edited by

London .. Thousand Oaks" New Delhi

Notes on Contributors ix

PART I INTRODUCTION TO THE EDITION

1 Introducing Qualitative Research

David Silverman

PART II OBSERVATION 9

2 Ethnography: relating the part to the whole 9

Isabelle Baszanger and Nicolas Dodier

3 Building bridges: the possibility of analytic dialogue between

ethnography, conversation analysis and Foucault 35
Gale Miller and Kathryn J. Fox

PART III TEXTS 56

4 Analysing documentary realities 56

Paul Atkinson and Amanda Coffey

5 Doing things with documents 76

Lindsay Prior

I) Internet communication as a tool for qualitative research 95

Annette N. Markham

PART IV INTERVIEWS 125

1 The "inside" and the "outside": finding realities in interviews 125

Jody Miller and Barry Glassner

8 The active interview 140

James A. Holstein and Jaber F. Gubrium

9 Membership categorization ami interview accounts 162

Carolyn Baker
CONTENTS

10 Focus group research

177
Sue Wilkinson

PART V TALK
200

11 Discourse analysis as a way of occurring talk 200

Jonathan Potter

12 Conversation analysis and institutional talk: analysing data 222

Paul Atkinson is Distinguished Research Professor in Sociology at Cardiff
John Heritage
University . He is Associate Director of the ESRC Research Centre on Social
and Economic Aspects of Genomics . His main research interests are the
PART VI VISUAL DATA 246 sociology of medical knowledge and the development of qualitative research
methods. His publications include: Ethnography: Principles in Practice (with
13 The conceptualization and analysiS of visual data 246 Martyn Hammersley) (Second Edition 1995, Routledge), The Clinical Experience
Michael Emmison (Second Edition 1997, Ashgate), The Ethnographic Imagination (1990, Routledge),
14 Analysing interaction: the visual and 266 Understanding Ethnographic Texts (1992, Sage), Medical Talk and Medical
Work (1995, Sage), Fighting Familiarity (with Sara Delamont) (1995, I-Iampton
Christian Heath Press), Making Sense of Qualitative Data (with Amanda Coffey) (1996, Sage),
Sociological Readings and Re-Readings (1996, Ashgate) and Interactionism (with
PART VII VALIDITY 283 William Housley) (2003, Sage) . Together with Sara Delamont he edits the
journal Qualitative Research. He was co-editor of The Handbook of Ethnography.
15 Reliability ami validity in research based on naturally occurring His ethnographic study of an international opera company is published as
social interaction 283 Everyday Arias.
Anssi Perakyla
Carolyn Baker was Associate Professor in the Graduate School of Education,
the University of Queensland, Brisbane, Australia. Her interests included
305
PART VIII THE WIDER COMMUNITY studies of talk and interaction in classrooms, meetings and other educational
16 Addressing social problems through qualitative research 305 settings. In her research on literacy she applied ethnomethodology and
conversation analysis to instances of 'talk around text' . An example of her
Michael Bloor
distinguished research output is her paper, with Jayne Keogh, 'Accounting
17 Using qualitative data and analysis: reflections on organizational for achievement in parent-teacher interviews', Human Studies, 18 (2/3), 1995.
research 325 Carolyn died in July 2003.
Gale Miller, Robert Dingwall and Elizabeth Murphy
Isabelle Baszanger is a sociologist at the CNRS (Centre National de la
Recherche Scientifique), Paris. Her main research interests are sociology of
PART IX POSTSCRIPT 342 medical work, sociology of pain and of the passage from curative to palliative
342
medicine, and interactionism . Her recent publications include Inventing
18 Who cares about 'experience'? Missing issues in qualitative research
pain medicine: From the laboratory to the clinic (1998, Rutgers University Press)
David Silverman and Quelle medecine voulons-nous? (with Martine Bungener and Anne Paillet)
(2002, La Dispute).
Appendix 368
Name Index 370 Michael Bloor is a medical sociologist with a personal chair in the School of
Subject Index 375 Social Sciences, Cardiff University. His recent publications include The
Sociology of HIV Transmission (1995, Sage) and (with Jane Frankland, Michelle
Thomas and Kate Robson) Focus Groups in Social Research (2001, Sage). He is

viii

NOTES ON CONTRIBUTORS
currently preparing for Sage a Dictionary of Qualitative Methods (with Fi�na
Wood) and The Sociology of Occupational Health and Safety: the case of the globalzsed
shipping industry (with Michelle Thomas).
Amanda Coffey is a senior lecturer in the School of Social Sciences, Cardiff
University. Her research interests include education in informal settings,
the role of (auto)biography in qualitative research and ethnographic repre­
sentations. She is currently co-editor of Sociological Research Online. Her
publications include The Ethnographic Self (1999, Sage) and Making Sense
of Qualitative Data (with Paul Atkinson) (1996, Sage). She was one of the co­
editors of The Handbook of Ethnography.
Robert Dingwall studied at Cambridge and Aberdeen, where he trained in
medical sociology before moving on to law and society research at Oxford.
Since 1990, he has been a professor at Nottingham, where he now directs
an interdisciplinary institute studying developments in biology and bio­
technology. He has published widely on professions, work, occupations,
interaction and qualitative methods.
Nicolas Dodier is a sociologist at the INSERM (Inshtut National de la Sante
et de la Recherche Medicale), and Professor at the Ecole des Hautes Etudes
en Sciences Sociales, Paris. His research interests include sociology of
technology, sociology of medicine and science, and theory of action. His recent
publications include L'expertise medicale: Essai de sociologie sur I'exercice du
jugement (1 993, Metailie), Les hommes et les machines: L� .
les societes technicisees (1995, Metailie), and Le(ons polltlques de I epldemle de slda
(2003, Editions de l'Ecole des Hautes Etudes en Sciences Sociales).
Michael Emmison is Reader in Sociology in the School of Social Science
at the University of Queensland, Brisbane, Australia. He works primarily in
the fields of language and interaction and material culture and is currently
researching the social organization of problem solving on a software helpline
and the impact of technology (telephone, email and online web counselli� g)
on troubles telling on a national children's helpline. He is the co-author (wIth
Tony Bennett and John Frow) of Accounting for Tastes: �ustralia� Every�ay
Cultures (1999, Cambridge University Press) and Researchmg the VIsual (WIth
Phillip Smith) (2000, Sage).
Kathryn J. Fox received her PhD in sociology from the University of California,
Berkeley, USA. She is now an associate professor of sociology at the University
of Vermont, USA, where she teaches courses on deviance and social con­
trol. She has completed qualitative studies of a midwestern punk scene in the
1980s, an AIDS prevention project in San Francisco and a cognitive therapy
programme for violent offenders in prison. Her new research project, wh�ch
is just beginning, is a study of truancy intervention and the social constructlOn
of truants.
x xi
(
NOTES ON C O NTRIBUTORS
Barry Glassner is Professor of Sociology at the University of Southern
California, USA, and the author or co-author of a dozen books, including Our
Studies, Ourselves (2003, Oxford), The Culture of Fear (2000, Basic) and Career
Crash (1994, Simon & Schuster). He has published papers in The American
Sociological Review, Social Problems, American Journal of Psychiatry and other
leading journals in the social sciences.
F. Gubrium is Chair and Professor in the Department of Sociology
at the University of Missouri, Columbia, USA. His research deals with the
narrative organization of personal identity, family, the life course, aging and
adaptations to illness. He is editor of the Journal of Aging Studies and author
or editor of more than twenty books, including Living and Dying at Murray
Manor (1975, St. Martin's Press), Caretakers (1979, Sage), qescribing Care (1982,
Oelgeschlager, Gunn and I-lain), Oldtimers and Alzheimer's (1986, JAI Press),
Out of Control (1992, Sage) and Speaking of Life (1993, Aldine de Gruyter).
Christian Heath is Professor of Work and Organisation at King's College
London. With members of the Work, Interaction and Technology Research
Group, he is currently undertaking studies of control centres, operating
theatres, medical consultations, newsrooms and museums and galleries. His
recent publications include Technology in Action (with Paul Luff) (2000,
Cambridge University Press).
John Heritage is Professor of Sociology, at UCLA, USA. He is the author of
Garfinkel and Ethnomethodology (1984, Polity), and The News Interview: Journalists
and Public Figures On the Air (with Steven Clayman) (2002, Cambridge
University Press) and the editor of Structures of Social Action (with Max
Atkinson) (1984, Cambridge University Press), Talk at Work (with Paul Drew)
0992, Cambridge University Press), and Practicing Medicine (with Douglas
Maynard) (2004, Cambridge University Press). He is currently working on
a range of topics in doctor-patient interaction, and on presidential press
conferences (with Steven Clayman).
James A. Holstein is Professor of Sociology in the Department of Social and
Cultural Sciences at Marquette University, Milwaukee, USA. He is the editor
of the journal Social Problems, and has authored or edited three-dozen books
including Court Ordered Insanity (1993, Aldine de Gruyter), Reconsidering Social
Constructionism (1993, Aldine de Gruyter), Challenges and Choices (2003, Aldine
de Gruyter) and Dispute Domains and Welfare Claims (1996, JAI Press). In
collaboration with Jay Gubrium, he has also published What is Family? (1 990,
M ayfield), The New Language of Qualitative Method (1997, Oxford University
Press), The Self We Live By (2000, Oxford University Press), Institutional Selves
(2001, Oxford University Press), Handbook of Interview Research (2002, Sage)
and Inner Lives and Social Worlds (2003, Oxford University Press).
Annette Markham is an assistant professor of communication studies
at the University of Illinois, Chicago, USA, where she teaches courses in
NOTES ON CONTRIBUTORS
T
NOTES ON CONTRIBU ORS

communication theory, interpretive research methods, and communication
technology. Her research focuses on the intersection of culture, technology
and identity from an ethnographic perspective. This research is well repre­
sented in her 1998 book Life Online: Researching Real Experience in Virtual Space
(AltaMira). Her current research explores the influence of computer-mediated
communication technologies on qualitative methods.
Gale Miller is Professor of Sociology, Department of Social and Cultural
Sciences, Marquette University, Milwaukee, USA. His research interests
include the study of language use in organizations, the sociology of troubles,
and the social construction of social problems. His interest in the paradox of
control is personal as well as intellectual, having just finished a term in
university administration.
Jody Miller is Associate Professor of Criminology and Criminal Justice at
the University of Missouri-St Louis, USA. Her research focuses on gender,
crime and victimization, particularly in the contexts of youth gangs, urban
communities and the commercial sex industry. Her monograph, One of the
Guys: Girls, Gangs and Gender, was published by Oxford University Press
in 2001 . She has published numerous articles and book chapters, including in
Criminology, Social Problems, Justice Quarterly, Journal of Research in Crime and
Delinquency and Journal of Contemporary Ethnography.
Elizabeth Murphy is Reader in Sociology at the University of Nottingham.
Her interests are in the sociology of health and illness, food, gender and
motherhood. She is conducting a study of the transition from child to adult
services for young people with learning difficulties. She is also interested in
the application of qualitative methods to policy-relevant areas. She trained at
the Universities of St Andrews, Southampton and the Open University.
Anssi Perakyi1i is Professor of Sociology at the University of Helsinki, Finland.
He received his PhD at the University of London in 1992. His research interests
include conversation analysis, medical interaction, psychotherapeutic
interaction and emotional communication. His publications include AIDS
Counselling (1995, Cambridge University Press) and numerous articles in
journals such as Sociology, Social Psychology Quarterly and Research on Language
and Social Interaction.
implication of studies of interaction for understanding cognition, and Focus
Group Practice (with Claudia Puchta) (2004, Sage), which analyses interaction
in market research focus groups.
Lindsay Prior is Professo r of Sociolog y at the University
of Wales, Cardiff.
He �s currently engaged in a number of projects that focus on lay and pro..
fesslOnal understandings of risk. These include studies of risk assessment
in cancer genetics, problems of genetics and insurance, lay attitudes to
vaccinations, and the use of anti-depressants in primary care. His mOst recent
book is Using Documents in Social Research (2003, Sage). .
David Silverman is Professor Emeritus of Sociology at Goldsmiths College,
UniverSity of London. His research interests are focused on professional-client
interaction, medicine and counselling, and qualitative research methods . He
is author of Doing Qualitative Research Second Edition (2004, Sage), Interpreting
Qualitative Data: Methods for Analysing Talk, Text and Interaction (Second Edition,
2001, Sage) and of Discourses of Counselling: HIV Counselling as Social Interaction
(1996, Sage). He edits the Sage Series called Introducing Qualitative Methods.
Sue Wilkinson is the Ruth Wynn Woodward Endowed Professor of Women's
Studies at Simon Fraser University, Vancouver, Canada. She is the founding
and current editor of the international journal Feminism [oj Psychology, and
author of over eighty publications in the areas of gender, sexuality and health.
Her current research interests are in breast cancer and threatened feminine
identities, and in conversation analytic approaches to data analysis.

Jonathan Potter is Professor of Discourse Analysis at Loughborough

University. He has studied racism, scientific argumentation and crowd
disorder. He is currently working on calls to a child protection helpline. His
most recent books include Representing Reality (1996, Sage), which attempts
to provide a systematic overview, integration and critique of constructionist
research in social psychology, postmodernism, rhetoric and ethnomethod­
ology, Talk and Cognition (with Hedwig te Molder) (Cambridge University
Press, in press), in which a range of different researchers consider the
xii
xiii
 David Silverman

The first edition of this book sought to provide a guide to the latest devel­
opments in qualitative research. This second edition offers a newly updated
introduction to cutting edge issues, written by leading scholars in our field.
Chapters from the first edition have been revised by their distinguished
authors. In addition, reflecting the changing face of qualitative research in
the past decade, four entirely new and exciting chapters appear. New to
this volume are chapters on visual data, focus groups, Internet data and the
applicability of qualitative research to organizational behaviour. To complete
these revisions, my concluding chapter on missing issues in qualitative research
has been specially written for this volume. Finally, to enhance the reader­
friendliness of this book, each chapter concludes with a set of annotated
recommended readings.
Like the first edition, this text aims to build on the success of my Interpreting
Qualitative Data (IQD; Silverman, 2001). Like that book, it was generated by a
number of assumptions set out below:
1 The centrality of the relationship between analytic perspectives and
methodological issues and the consequent requirement to go beyond a
purely 'cookbook' version of research methods.
2 The need to broaden our conception of qualitative research beyond issues
of subjective 'meaning' and towards issues of language, representation and
social organization.
3 The desire to search for ways of building links between social science
traditions rather than dwelling in 'armed camps' fighting internal battles.
4 The belief that a social science, which takes seriously the attempt to sort fact
from fancy, remains a valid enterprise.
DAVID SILVERMAN s INTRODUCTION
INTR O D U CING #
QUALITA TI VE RESEA CH

5 The assumption that we no longer need to regard qualitative research as student audience. This means that the presentation is didactic but not
provisional or never based on initial hypotheses. This is because qualitative 'cookbook' in style.
studies have already assembled a usable, cumulative body of knowledge. The particular contribution of this reader lies in its assembly of a very well­
6 The commitment to a dialogue between social science and the community known international team of researchers who share my commitment to
based on a recognition of their different starting points rather than upon a rigorous, analytically derived, but non-polarized qualitative research. Eight
facile acceptance of topics defined by what are taken to be 'social problems'. US researchers join eight from the UK, two from France and Australia and
one from Finland and Canada. While the majority of the contributors are
ted within sociologists, the disciplines of social psychology, criminology and educational
Each of these assumptions is, implicitly or explicitly, highly contes e, becau se such studies are also represented. In any event, I believe that all contributors have
This is largely , I believ
suc �eeded in making their presentations accessible to a multidisciplinary
contemporary qualitative resear ch.
research has become a terrain on which diverse schools of social theory have
try to audience. Rather than denying their own analytic position in favour of some
fought their mock battles. Ultimately, the assumptions set outayhere ces wo?lly centre ground, these authors have clearly set out the assumptions from
move the terrain of our field towards an analysis of the everyd isresour whIch they proceed while remaining open to the diverse interests of their
which we use in making our observations. This point, which r of thisit implic
readers. Each has written a chapter which reflects on the analysis of each of
in many of these contributions, is set out in detail in the final chapte the kinds of data discussed in IQD: observations, texts, talk, visual data and
book.
Of course, avoiding such battles, in the context of a commitment to a interviews. Following IQD, each author uses particular examples of data
r releva nt analysis to advance analytic arguments.
cumulative social science, is far more likely to make our trade appea s, with The two chapters on observational methods seek to rescue observational
to the wider community. As we look outwards rather than inward ue between work from the pitfalls of mere 'description' and lazy coding and towards
confidence rather than despair, the way is open for a fruitful dialog s. exciting methodological and analytic directions for observational research.
social scientists, organizations, professionals and community group to the wider In Chapter 2, Isabelle Baszanger and Nicolas Dodier begin with the need
Moreover, it is worth noting that we present ourselves not only Researc h
community but also to the students we teach. Both Doing Qualitative
to ground research in field observations. The question they then raise is how
(Silverman, Second Edition 2004) and IQD derive from thirty years under­ of teachin g the ethnographer actually goes about relating partial observations to broader
generalizations about the 'whole'. Baszanger and Dodier show how ethnog­
methodology courses and supervising research projects at both the wisdom raphy has been dominated by traditions which seek to integrate observations
graduate and graduate levels. That experience has reinforcedpracti
of the old maxim that true learning is based upon doing. In ce, this either by an appeal to the concept of 'subculture' or by the understanding
studen ts are or writing of the individual author. Rejecting such appeals to 'culture' or 'the
means that I approach taught courses as workshops in which This means self', they depict a 'combinative ethnography' which seeks to generalize by
given skills to analyse data and so to learn the craft of our trade. h data applying the comparative method to groups of situations or activities
that assessments of students' progress are properly done throug invited to collected in the ethnographic 'casebook'.
exercises rather than the conventional essay in which students are
offer wooden accounts of what other people have written. In Chapter 3, Gale Miller and Kathryn Fox show how cumulative observa­
It follows that I have little time for the conventional trajectory of thegather PhD ti�n can be combined with analytic vitality. In this chapter, 'Building Bridges',
re', MIller and Fox raise the possibility of dialogue between ethnography,
in which students spend their first year 'reviewing the literatu they can conve�sation analysis and Foucault. Beginning with the focus on naturally
data in the second year and then panic in the third year about howin the first occurrmg data used by discursively oriented ethnographers, Miller and Fox
analyse their data. Instead, my students begin their data analysis 'cracked' point to what each of these three traditions have in common and to how they
year - sometimes in the first week. In that way, they may well have their can provoke a set of fascinating research questions for the ethnographer. They
the basic problem in their research in that first year and so can spend atic tasks then show how these questions can be addressed in the single case study as
remaining years pursuing the worthy but relatively non-pr oblem
method. well as in comparative or longitudinal studies.
of ploughing through their data following an already-establishedwho
Like IQD, my hope is that this book will be used by studen ts are not
pinnin gs
Par.t III on 'texts' follows Miller and Fox's call for building bridges by
showmg how ethnographic reading of texts can fruitfully work with a diverse
yet familiar with the approaches involved, their theoretical under
and their research practice. In IQD, student exercises were design ed to allow
d­
set of analytic traditions. Paul Atkinson and Amanda Coffey apply theories
from the literary theory of narrative and genre to the documents through
readers to test their understanding of each chapter. In this book, worke more which organizations represent themselves and the records and documentary
through examples of research studies make the argum ents much
collection data they accumulate. Taking the example of 'audit', they show how we can
accessible. Moreover, the chapters are not written in standard editedssible to a fruitfully analyse financial statements produced by accountants and accounts
style as chapters addressed to the contributors' peers but inacce
3
2

DAVID SILVERMAN " INTRODUCTION
of their work by university departments. They also remind us of the 'audit
trail' as documents refer to other documents. Following Atkinson and Coffey,
we are given the tools to explicate systematically how texts are organized
through the concepts of 'authorship', 'readership', 'intertextuality' and
'rhetoric' .
In Lindsay Prior's chapter on texts, we move from literary theory to theories
of discourse. However, unlike the stultifying theoretical level of some
introductions to this topiC, Prior has written a delightful, accessible chapter
which shows, in practice, what it is like to 'do things with documents'.
Avoiding references to a knowing 'subject', Prior shows us how we �an
instead focus on the ways in which a text instructs us to see the world. Usmg
examples as diverse as a statistical summary of 'ca� ses of death' and a
psychiatric interview, he reveals a thought-provokmg toolbox that we
can use when working with textual material.
In the twenty-first century, however, conventional documents are not the
only textual material that circulate in the world. The Internet is now perhaps
the prime site where words and pictures cir.culate. Annette Marl�ham' s
new chapter develops this insight and, in so domg, offers reader� an mvalu­
able guide to interpreting such data. Markham shows the lmp �rtance
of distinguishing three ways in which the Internet works: as a medIUm .of
communication; as a network of computers; and as a context for sOClal
interaction. Using illuminating examples of Internet data, Markham dem?n-·
strates how researchers can use the Internet either as a means of conductmg
conventional interview or focus group studies (albeit with different time
constraints) or as a way of studying how participants themselves constitute
meaning in naturally occurring websites such as chatrooms. Follow.i�g this
latter option, we learn, as in the other chapters on texts, how partlclpants
actively construct meaning.
.
This idea of the 'active' reader is carried over into Part IV on mtervlews
.
and
focus groups. All four chapters in this section remind us that both respondents
and social scientists actively construct meaning in each other's talk. Jody
Miller and Barry Glassner address the issue of finding 'reality' in inte�view
accounts. As I argue in IQD, the desire of many researchers to treat mter­
view data as more or less straightforward 'pictures' of an external reality can
fail to understand how that 'reality' is being represented in words. Miller
and Glassner set out a position which seeks to move beyond this argument
about the 'inside' and the 'outside' of interview accounts. Using their own
research on adolescents' social worlds, they argue that interview accounts
may fruitfully be treated as situated elements in social worlds, drawing upon
and revising and reframing the cultural stories available in those �orlds. For
Miller and Glassner, the focus of interview research should be flxed upon
what stories are told and how and where they are produced.
In their chapter, James Holstein and Jaber Gubrium show us how a focus
on story and narrative structure demands that we recognize that both
interview data and interview analysis are active occasions in which meanings
are produced. This means that we ought to view research 'subjects' not as
4 5
r
INTR O D U CI N G QUALITATIVE RESEARCH
stable entities but as actively constructed through their answers. Indeed, in
Holstein and Gubrium's telling phrase, both interviewee and interviewer are
'practitioners of everyday life'. Using examples from their research on nursing
home residents and on carers of elderly family members, they invite us to
locate the interpretive practices which generate the 'hows' and the 'whats' of
experience as aspects of reality that are constructed in collaboration with the
interviewer to produce a 'narrative drama'.
The final chapter on interview data is by Carolyn Baker. In common with
Holstein and Gubrium, Baker treats interview talk as social action inrwhich
all parties draw upon their cultural knowledge in doing their accounting
work. Baker's particular contribution is to show how interview data may be
analysed in terms of the categories that participants use and how those
categories are routinely attached to particular kinds of iiictivity. Using this
form of Sacks's 'membership categorization analysis' (see also Part IX), Baker
shows how we can describe the interpretive work present in data taken
from parent-teacher interviews and research interviews with teenagers and
the Chair of a school welfare committee. Like the previous two chapters,
Baker's appeals to the 'cultural logics' drawn upon by members in accounting
for themselves and assembling a social world which is 'recognizably familiar,
orderly and moral'.
Sue Wilkinson's chapter on focus groups carries forward Baker's focus on
how we construct the social world with our respondents. Using illuminating
extracts from her own data, Wilkinson reveals the complicated interpretive
activities between members of focus groups as they try to make sense of each
other (and the researcher ). This close attention to the details of short data
extracts is contrasted with how most focus group (and interview) research is
usually conducted. Wilkinson's concern with theoretically driven, detailed
data analysis stands apart from the dominant tendency to treat focus group
talk as a straightforward means of accessing some independent 'reality'.
Above all, Wilkinson shows us that content analysis and a concentration on
�he mechanics of how to run a focus group are no substitute for theoretically
mformed and detailed data analysis of talk-in-action. Like all the contributors
to this volume, Wilkinson underlines the fact that we must never overlook
the active interpretive skills of our research subjects.
Part V is concerned with audio data. Jonathan Potter discusses discourse
analysis (DA) as a way of analysing naturally occurring talk. Potter shows
the manner in which DA allows us to address how versions of reality are
produced to seem objective and separate from the speaker. Using examples
drawn from television interviews with Princess Diana and Salman Rushdie
and a newspaper report of a psychiatrist's comment, he demonstrates
how we can analyse the ways in which speakers disavow a 'stake' in their
actions.
In its focus on how reality is locally constructed, DA shares many concerns
with conversation analysis (CA). John Heritage'S chapter presents an acces­
sible introduction to how conversation analytic methods can be used in the
analysis of institutional talk. After a brief review of the main features of such
DAVID SILVERMAN INTRODUCTION
talk, Heritage devotes the rest of his chapter to an illuminating analysis of
a short telephone conversation between a school empl,oyee and th.e mo:her of
a child who may be a truant. He shows how, using CA, we can Iden:Ify . the
overall structural organization of the phone call, . its seq�ence orgamzatl?n,
turn design, the lexical choices of speakers and mteractIOnal. asym.metnes.
Finally, Heritage demonstrates how each of these elements fIts mSIde each
other - 'rather like a Russian doll', as he puts it.
The elegance of Heritage's account of instituti� nal talk is .m�tched �y the
two chapters in the next part on visual data. LIke Sue �Ilkmson (m her
chapter on focus groups), Michael Emmison argues that vIsual res�arch�rs
have worked with inadequate theories. For instance, most tend to Identify
visual data with such artefacts as photographs and, to a lesser extent, cartoons
and advertisements. Although such work can be interesting, it is, in a sense,
two dimensional. If we recognise that the visual is also spatial, a whole
new set of three-dimensional objects emerge. By looking at how people use
objects in the world around them (from streetmaps to the layout of a room),
we can study the material embodiment of culture.
Christian Heath's discussion of the analysis of face-to-face interaction
through video shows one way of looking at three-dimensional d�ta in fine
detail. Beginning with a clear account of CA's focus on sequential organ­
ization, Heath shows how CA can be used to study visual conduct and how
the physical properties of human environments are made relevant within the
course of social interaction. Like Heritage, Heath uses an extended example.
In a medical consultation, a patient's movements serve to focus the doctor's
attention on a particular aspect of her account of her symptoms. The example
also shows that, while the visual aspect of conduct is not organized on a turn­
by-turn basis, as Heath puts it: 'the sequential re�ations �et"'.'ee� visual
and vocal actions remain a critical property of theIr orgamzatIon . Heath
concludes by showing the relevance of these insights to studies of the work­
place, including human-computer interaction.
The final four chapters of this book, by Perakyla, Bloor, Miller, Dingwall
and Murphy and myself, move on to broader themes about the credibili�y a.nd
wider impact of qualitative research. Anssi Perakyla discusses how qualItative
research can seek to offer reliable and valid descriptions. Following Heritage's
chapter, Perakyla illustrates his argument with CA research on institutional
interaction. He shows how good transcripts of audio-recorded interactions
can maintain the reliability of the data. However, Perakyla also shows how
we can accommodate the fact that tapes do not necessarily include all aspects
of social interaction and addresses such 'nitty gritty' questions as the selection
of what to record, the technical quality of recordings and the adequacy of
transcripts. Finally, validity questions are discussed in terms of conve�tio�al
'deviant case analysis' as well as specifically CA methods, such as vahdatIon
through 'next turn'. Overall, Perakyla is right to claim that his chapter is the
first systematic attempt to discuss such matters in relation to �A. At th� sa�e
time, his discussion has a much broader relevance to all senous qualItatlve
research.
6 7
INTRODUCING QUALITA TIVE RESEARCH
�ichael Bloor's chapter also deals with a topic that concerns most quali­
tatIve researchers: the ability of our research to contribute to addressing social
problems. Bloor argues that our focus on everyday activities makes it
particularly relevant in helping practitioners to think about their working
practices. He demonstrates his argument by detailed discussions of case
studies which he conducted of male prostitutes in Glasgow and of eight
therapeutic communities. Both sets of studies illustrate Bloor's point about
the ways in which rigorous qualitative research can have relevance for service
pro,"ision, even if, at least in the UK, it is unlikely to have much impact upon
pohcy debates at the governmental level. Finally, Bloor reviews (and rejects)
the argument that social scientists should not be practitioners' helpers.
Bloor's focus on how professionals can make use of qualitative research
is complemented by Miller, Dingwall and Murphy'S' chapter. Like Bloor,
they are concerned with the wider community. However, their attention is
on the variety of 'stakeholders' in the organizations that dominate our lives.
Economists and management consultants hold centre stage in this arena and
qualitative research receives little attention. Yet the latter's ability to reveal
or�ani� atio�al processes suggests that we have much to offer to managers.
. g examples of studies of
USIng Illummatm both private corporations and
publ.ic � gencies, Miller, Dingwall and Murphy establish precisely what
quahtahve research, with its flexible research designs, can offer organizations.
Organizational complexities can be recognized and, as a result, new ways of
reframing organizational problems can be posited.
Not all of the contributors to this volume are in agreement about every
issue. We particularly see this within Parts II and V, where contrasting views
of each kind of data analysis are advanced. None the less, I believe that the
contributors to this volume share enough in common to make this a coherent
volume. Many of my contributors, I suspect, would agree with most of the six
points at the start of this chapter. With more certainty, I would claim that we
share a fairly common sense of what constitutes 'good' qualitative research.
For instance, even though we come from different intellectual traditions,
I would be surprised if we were to have any fundamental disagreement about,
say, the assessment of an article submitted to us for refereeing.
This common sense of what we are 'looking for' derives, I believe, from an
attention to the mundane properties of everyday description. Therefore, this
� olume concludes with a postscript, drawing upon the work of Harvey Sacks,
In which I sketch out these properties and their consequences for qualitative
research. I thank Geraldine Leydon, Jay Gubrium and Judith Green for their
comments on an earlier draft of this chapter.
I want to conclude this introduction by mentioning an absent friend.
Carolyn Baker had agreed to revise her brilliant chapter on interviews for this
volume. Tragically, a serious illness prevented her fulfilling this commitment.
Sadly, Carolyn died a few days before I wrote this introduction. She will be
s?rely missed for both her intellectual brilliance and personal qualities. In the
CIrcumstances, I have limited myself to some minimal updating of her chapter
for this volume.
DAVID SILVERMAN " INTRODUCTION

As always, my thanks are also due to Gilly for putting up with me and to my
friends at the Nursery End for giving me summers I can look forward to.

Silverman, D. (2000) Doing Qualitative Research. Londo n: Sage.

for Analysmg Talk, 1 ext and
, r "

Silverman, D , (2001) Interpreting Qualitative Data: Methods

Interaction, Second Editio n. Lond on: Sage.

Relating the the whole

Isabelle Baszanger and Nicolas Dodier

By the beginning of the twentieth century, the anthropological tradition,

primarily through the influential work of Bronislaw Malinowski, Edward
Evans-Pritchard and Margaret Mead, had conferred an abiding legitimacy
on field observations integrated into a 'cultural whole'. The subsequent
crisis in this model corresponds to two lines of questioning. First, ethno­
methodological studies undermined the conventional view by revealing the
interpretations and negotiations needed to decontextualize observation
situations at all junctures of fieldwork. Whether that work is ethnographic
or statistical, it involves invisible operations that do not generally appear
in social science texts. Second, analysis of how field notes are constructed
forced a revision of the traditional views of the anthropologist in the field
and served as a basis for a very critical reassessment of the authority of the
ethnographer.
In parallel, within the field of sociology, greater emphasis was being
placed on ethnographic approaches, primarily in terms of the importance
given to direct, in situ observation of concrete sequences of activities. This
was encouraged by new developments in the theory of action including exten­
sions of the interactionist tradition, the many studies focusing on scientific
activity, and work carried out in French pragmatic sociology looking at
different regimes of action .
Since that time, new conceptions of ethnography have emerged. They
reassert the value of fieldwork, but focus more on demonstrating the rela­
tionship between forms of heterogeneous action rather than trying to identify
a culture as a whole. This chapter will take stock of these new developments
in ethnography, the underlying conception of the field and the way in which
they deal with the fact of human plurality (Arendt, 1995).

8

ISABELL E BASZAN GER AND NICOLAS DO DIER
w OBSERVATION
We will start with the question of the totalization of ethnographic data,
i.e. how to integrate the series of data collected in the field into a single whole?
We suggest distinguishing between three kinds of ethnography:
1 Integrative ethnography: following the anthropological tradition, this
idual s.
constructs units of collective belonging for indiv first-person narrative
2 Narrative ethnography: by contrast, this offers readers a
of events for each different field.
3 Combinative ethnography: by working simultaneously in fy
different fields, this
the different forms
brings together a casebook that can be used to identi
of action in which people may engage, along with the possib le combinations
between them.
As the above list suggests, totalization has a methodological dimension.
This involves an examination of the operations whereby a totality is built up
in the course of an ethnographic study, into which each concrete item is then,
finally, fitted. These operations are both external (note taking, building up a
series of observations, codings, etc.) and internal (the transformation, during
the work, of the researcher's own way of apprehending the world). At the
same time, it is also a theoretical question, since the relationship between
individual cases and a totality directly involves the status ascribed to the
references to the latter (e.g. collective belonging, individual stories, forms
of action) in interpreting individual behaviour or actions.
In this chapter, we show how combinative ethnography has helped find
a way out of the crisis of totalization in its 'integrative' version, and also how
the question of the anchoring of the wholes thus constituted is brought into
new focus by initiating reflection around the mode of otherness in which
persons appear in each of the three forms of ethnography listed above.
AN I N ITIAL CHARACTERIZATION OF ETH N O G RAPHIC
R ESEARCH
Ethnographic studies are carried out to satisfy three simultaneous
requirements associated with the study of human activities:
1 the need for an empirical approach;
2 the need to remain open to elements that cannot be codified at the time
of
the study;
3 a concern for grounding the phenomena observed in the field.
Each requirement is briefly discussed below.
The need for an empirical approach
d cannot
This first need is dictated by the fact that the phenomena studie
btedly what
be deduced but require empirical observation. This is undou
Durkheim really meant by his well-known injunction to 'treat social facts
10
I
ETH NOGRAPHY
as things',. meaning not so mu�h that sociology should be conducted along
th� same lmes as the natural SCIences, but as a way of distinguishing it from
phIlosophy and the introspection that takes place upstream of an empirical
approach. 1 In the current debate over the resources people mobilize to
understand the �orl� and to make reference to it, this is the major difference
between the SOCIal SCIences, on the one hand, and the philosophy of language,
phenomenology and hermeneutics, on the other.
The need to remain open
Beyond any methodical planning of observations, the fieldworker must remain
open in order to discover the elements making up the markers and the tools
th�t people mobilize in their interactions with others and, more generally,
�Ith the world. By markers, we mean representations of the world, or norma­
tIve expectations, but also the linguistic and para-linguistic resources that
are displ�yed in contact with the environment (Bessy and Chateauraynaud,
1995; Theven?t, �994). �he objective here is to distinguish between openness
to new data (m SItu studIes) and its opposite, as when individual activities are
�tudied according to strict schedules and on the basis of previously defined
�tems and rules (a �riori codified studies). This second approach is intrinsically
mcapable of revealmg the unexpected elements that come to light as a study
progresses. In methodological terms, a study can be described as in situ if
it allows each subject to behave in an endogenous manner: that is, one that is
not influenced by the study arrangements.
�re many reasons for not 'aligning' the subjects of a study in com­
. ThereWIth
plIance the study arrangements, just as there are many theories calling
for recourse to ethnographic studies: discovery of other cultures that cannot
be under�tood in the light of pre-existing knowledge (anthropological tradition),
the contmgency of continually negotiated human activities (interactionist
tradition) and observation of how people handle the contingencies of a given
situation (ethnomethodology), and so on.
. This p�incipl� of openness to what cannot a priori be pre-codified results
111 the baSIC ten� lOn u derlying in situ studies. The flexibility required by this
�
openness conflIcts WIth the need to maintain at least a minimum of method
in the conduct of the study; that is, a certain guide for the behaviour of
both the f�eldworker and the people observed, depending on the plan of the
study. ThIS duality is an implicit part of the general situation of the in situ
fieldworker.
The tension is primarily epistemological. The principle of non-alignment
of the people observed �oes not sit easily with the principle of planning that
has governed the expenmental sciences since the idea of scientific 'repro­
duct�on' or 'reproducibility' was elevated to the rank of a major, normative
reqmrement of scientific research (Licoppe, 1996). Social scientists who wish
none the Ie.ss: �o continue openly to observe the endogenous development of
human actIVItIes approach this problem in a number of different ways. Some
seek to conform as closely as possible to the requirements of experimental
11
ISABElLE BASZANGER AN D N I COLAS D O D I ER OBSERVATION
ETHNOGRAP W j
0)

reproducibility. Even if they allow open activity sequences to take place, they Tl:is does not exclude a secon d step, in which a series ethno graph
try as much as possible to standardize the time intervals of these sequenc�s studIes can serve as sources for defining universal, humanofpheno mena
ic
and record this activity by automatic means (tape recorder or VCR used m the t�ue sens � . This is, for example, the position of Levi-Strauss: ethnograph , in
conversation analysis see Parts V and VI of this volume). studIes provlde elements for ethnological texts that study societ ic
Others oppose this requirement of alignment and even the whole idea of the other. On this basis, starting from a systematic comparison betwe ies one after
observation corresponding to the current canons of science. They insist on an societies, anthropological work attempts to arrive at a theory of the struct en
approach that is opposed to any type of planning, leaving the study completely of the human spirit. ure
open to the uncertainties of the field. Still others recognize the need for
some sort of compromise between method and openness to situations, and . In .in
tmgUl
situ studies, this reference to field experience
never
shes ethnographic studies from other observation methods theles s dis­
see ethnographic tensions as a more extreme but, ultimately, a quite banal that
grounded in a specific field (conversation analysis, situated cognitio'n not are
example of the sort of negotiation that is omnipresent in science (e.g. scientists' ethnomethodo]ogy). This raises questions about fieldwork: what is the status and
negotiation between the need to follow standard rules, which in any case of this 'specific' context in which the study takes place? How is it described?
always demand local interpretations and adjustments, and the concrete course How is this framework delineated, since it is not a here-�nd-no
of any scientific endeavour)? nor a situation in wh�ch humankind as a whole is characterized w situation,
throug h the
It is worth noting that this duality underlying the ethnographer's work fundamental properties of every one of its activities?
also has moral implications beyond the epistemological dimension. To Thes� q� es �ions are the focal point of any analysis
satisfy this principle of openness, which is deliberately taken quite far, the generalIzatIOn m ethnography, particularly in terms of howofa the process of
cultur al whole
ethnographer must graft his/her study onto pre-existing systems of activity. is depicted. This is what is called here the process of 'totalization', an opera
As opposed to the researcher, who channels subject matter into the laboratory, whereby the ethnographer integrates the different observation sequences tion into
the ethnographer leaves the laboratory and tries to make his/her data gather­ a global referential framework.
ing compatible with the study population's other commitments. By definition,
ethnographic study design is a hybrid approach in which the field worker is
present in two agencies, as data gatherer and as a person involved in activities
directed towards other objectives.3
The ethnographic tradition has long considered that it is possible to integrate
sequences of ethnographic observations by relating them to a
Grounding observed ....n"'nnrrtl"'YI::J in the field
a global reference which encompasses these observations andcultur within
al whole:
A study becomes ethnographic when the fieldworker is careful to connect the the different data throw light on each other. This vision of an integrwhich
facts that s/he observes with the specific features of the backdrop against which e.thnogra�hy has been developed in social and cultural anthropology, ative par­
these facts occur, which are linked to historical and cultural contingencies. ticularly m the study of non-Western societies, but also in similar studie
Not all in situ studies are field studies. carried out in Western countries. It also involves the most culturally orienteds
Distinctions can be made between different sorts of empirical study carried part of the interactionist tradition: that is, the study of microcultures or
out in the social sciences. Some attempt to universalize, i.e. are formal in subcultures, and, more generally, all references in the interactionist tradition
nature, while others, resolutely grounded in a specific context, can be con­ to the existence of communities of peop Ie sharing the same rules and the same
sidered as ethnographic or field - studies. Formal studies dissociate collected understanding of the world (e.g. deviant communitie s).
data from any context in order to access a universal, human level from the
outset. This is the approach characteristic of the philosophy of language, of . I�tegrat
dlstmc
ive ethnography propo ses a monographic totalization that is
t from statistical totalization or summation. In general, the
Austin's pragmatics, of phenomenological analyses or ordinary language not meet the requirement of openness, and is therefore excludedlatter from
does
analysis based on analysis of conversations. classification of 'in situ' study methods.4 our
It is also characteristic of the 'nomothetical' approach that uses empirical A. nu�ber of n:ethods have been proposed to achieve this mono graphic
observation to demonstrate consistencies between facts and to formulate totahzatlO n. �lrst,
. It can result
from the fieldworker's reflect
general laws. Nevertheless, the ethnographic study is not only empirical achIeves an mtegrated vision of his/h er subjective experions,
. whereby s/he
ience.
or only 'open'. It is, like history, embedded in a field that is limited in time meaning behind calls for empathy with the people encounteredThis : the
is the
and space (Ricoeur, 1984-1988). Returning to a concept that Darbo-Peschanski worker tries to immerse him/herself in the field conditions field­
(1987) applies to the studies of Herodotus, it is a 'science of the particular' to �he point of view of the others, seen as 'natives': that is, peopl
and gain access
a S ImIlar cultural persp
and describes itself as such. . e who share
ective, one that differs from the perspective of the
12
13
ISABELLE BASZANGER AND NICOLAS D O D IER OBSERVATION
newly arrived fieldworker. As Clifford noted, a professional anthropologist
is supposed successfully to 'infiltrate the expressional universe of the other'
(1983: 100). By understanding the other through an empathetic relationship,
the fieldworker would be able to reconstruct this other's point of view, and
therefore culture (or the contents of this other's collective consciousness).
The assumption of empathy as the process through which the point of view
of the other becomes transparent to the fieldworker is vulnerable to criti­
cisms arising from hermeneutic interpretation of texts and actions (Gadamer,
1975; Ricoeur, 1991). The act of gaining access to the point of view of the
other always implies an initial period of questioning, which is itself embedded
in a certain tradition, that of the interpreter - s/he is always caught in the
'hermeneutic circle' of the initial questioning of the text (or action), and of its
transformation, in return, as a result of this encounter.
Although certain relationships or certain moments can be better qualified
as empathetic, in the sense of a type of harmony between persons, we cannot
therefore conclude that the point of view of the other will be conveyed in
total transparency or that it can be expressed in words. Any interpretive act
is influenced, consciously or not, by the tradition to which the interpreter
belongs. Levi-Strauss (1963) proposes a variation on this process by con­
sidering the moment the experience is integrated into a whole not as a moment
of access to the point of view of the other, but as a moment when the entire
set of results experienced and memorized by the fieldworker crystallizes
into a single, unified experience. That event takes place at the conclusion of
the fieldworker's ethnographic apprenticeship in the society studied. The
ethnographer's field experience 'represents a crucial moment in his/her
education, prior to which he may have accumulated dissociated knowledge
that might never integrate into a holistic experience; only after this moment
will this knowledge "take definitive form" and suddenly acquire a meaning
that it previously lacked' (Levi-Strauss, 1963). In other words, what we are
dealing with here is a genuine 'internal revolution' (Levi-Strauss, 1963).
By treating participant observation as a method rather than 'a clinical talent'
resulting from an empathetic stance,S a number of authors have helped
to provide a new explanation for the position of the ethnographer, which
becomes one of his/her own methodological tools. Just like the ethnographer
in remote societies, the observer has to accept a separation from his/her
familiar universe, not only in order to be physically present in the new
environment, but also in order to achieve personal proximity.6
The observer has to enter into the group and find the right distance
between him/herself and the group. There is a close relationship here between
the observer's presentation of him/herself (to enter the field and throughout
the study), and the place accorded to the observer by the other.
While it is paramount for a fieldworker to be attentive to the expectations
and role projections of the people being observed/ this is less in order to
achieve an empathetic attitude than because the interrelations themselves
and, ultimately, the fieldworker and the work done on his/her experience
are the preferential instruments of observation.
14
ETHNOG RAPHY
There is nothing romantic or intuitive here (Piette, 1996: 68-72) . This
is conscious work on the part of the observer, who has to contro l his/he r
emot�onal reaction to what is observed and also develop a finely tuned
spectIOn to understand fully the process of transformation which intro­
undergoes by �eing �onstantly present in the field. Hence, understandis/he ng a
cultural whole IS achIeved through this reciprocating motion of the observ
an� the pheno:nena that s/he is observing, or, as Fox puts it, the process 'by er
whIch a partlc. Ipant observer gradually makes organized sense out of what
he sees, hears and becomes a part of it' (1974: 230).
The observer establishes a sort of parallel between what s/he feels and what
the people observed feel, or the phases they pass through. S/he uses a form
of introspection to reveal how s/he develops new attitudes or borrows
new roles and what that 'does for him/her'. In this way, the observ er has
fleeting insights into the possible functions and me�nings for the people
observed, whIch . s/he then
tries to verify in the field, at which point s/he either
recognizes their validity or rejects them. s
The duration of the observation enables the fieldworker a sense to
immerse him/herself in the subject being observed, but this in
process of socialization than a direct access to the point of viewisofcloser the
to a
other.
In order to achieve a comprehensive understanding of a group, the fieldworker
has to work his/her way through the dense fabric of the culture observ ed, in
orde: to arri�e l�ter �t an objective und�rstanding �d, hence, a monographic
. �
totahzc atIon. ThIS bnngs us back to a slIghtly modIfIed form of Levi-Strauss's
position.
Geertz (1973) also distances himself from the empathetic schema and re­
integrates the concept of culture in a hermeneutic process: activities can be
read like texts, as far as both the actors themselves and the fieldworke r are
c oncerned; the concept of 'culture' is also a tool for the actors, who use
to interpret their reciprocal behaviour. It is the discovery of the hermeneuticit
role of the concept of culture for all individuals in their daily relationships
based not on a representation of lived experience or on their point of view,
but on a description of their oral or written production -that allows us to relate
a seq�ence of specific scenes to a culture. In any case (empathy, the integra tive
expenence of the ethnographer, the participant observation of the sociologist,
the �ermeneutic appro�ch), access to the major features of the culture being
studIed, as can be seen 111 all ethnography manuals, implies
go beyond a disparate set of ethnographic observations using and
methods to
discov er an
integrated culture which is different from other cultures.IO
Because of its capacity to satisfy the need for concrete
and at the same time produce a discourse taking in collectfacts ive
in the study
whole
approach has for a long time exerted a great deal of fascination over thes,social this
�c �ences. However, it does not stand up very well to two criticisms. First,
�t IS .o�ly valid if we are dealing with so-called 'mechanical' solidarity between
mdIv iduals (Durkheim): that is, a SOciety or group in which people are
assumed to share the same elements of the collect
arise if the coordination between human activitive consciousness. Difficulties
ies conforms to other types
15
 ETHNOGFlAP � Y
ISABELLE BASZANGEFI AND N I COLAS DODIER 0 OBSERVA TION

of logic. In contemporary worlds, one now has to take into account the fact
that several possible references can coexist despite their contradictions,
sometimes within the same person, and that they can slot into the normative
guidelines for action depending on the constraints of the actual situation
(Boltanski and Thevenot, 1991).
The very notion of society becomes problematical when solidarity between
people is established along socio-technical � etworks in. which �nd�viduals
coordinate their activities step by step accordmg to functIOnal ObjectIves and
without reference to a single common framework (Dodier, 1995a). Moreover,
at the methodological levet the moment at which data are integrated into
a whole occurs at an unknown, almost mysterious point of the process.
Some ethnographers cultivate this sense of mystery by affirming their
lack of interest in any account of methods. But others take a more ambiguous
stand: monographic totalization may conceal implicit statistical totalization
performed, as it were, behind the scenes. We see this particularly in the rhetoric
of cultural ethnographic studies, which very easily use frequency markers
when describing behaviour (often, sometimes, from time to time, always,
traits. The ethnographer is seen as an author and due emphasis is given to
the profoundly personal nature of his/her account (Clifford and Marcus, 1986).
Another channel consists of seeing the fieldwork itself not as the hidden
face of ethnography, perhaps reported in a personal diary, but as the actual
material of the ethnographic text. This text is no longer the 'picture' of a culture
or � soci �ty revealed to the ethnographer at the end of a learning process by
whIch, fmally, s/he is able to see it as a whole, but the 'account' of events
confronting the ethnographer as the enquiry progresses. The narrative is now
s�en as .integrating these events (Ricoeur, 1991); it preserves their temporal
dImensIOn and does not banish the ethnographer from his/her text - quite the
opposite, in fact.
Such narrativ� ethnography can take the form of an approach that we might
call hyper-re�lexIve, more preoccupied in fact with questiol}ing and reporting
on the operatIOns performed by the ethnographer in his/her attempt, through
concepts such as 'culture' or 'society', to confer some meaning on activities,
than really acknowledging the existence of the other (Moerman, 1972). The
��c?unter betwee� et��ologist and study population is viewed as dialoguea

etc.) without making more than a token attempt to justify them. These prob­ InItlated between mdividuals who themselves belong to different collective
lems have been taken very seriously by some historians of ethnography wholes. The concept of 'culture' is not abandoned, but the ethnographer does
and by certain ethnographers, to the point of casting doubt on the tradition not t�y so m�ch to acknowledge an 'other culture' as to reveal the dialogue
of totalization presented more or less as a foregone conclusion in ethnology that IS esta bhshed between different cultures during the fieldwork (Dwyer'
and field sociology. 1979).
Fin�lly, the narrative approach in ethnography is influenced by psycho­
�nalysis. (Favret-Saada, 1980; Favret-Saada and Contreras, 1981).11 The study
NARRATIVE I NTEGRATIO N
IS descnbed as a process profoundly linked to the individual history of the
ethnographer. The text may be a history of the events occurring in the course
the famous of the fieldwork, i.e. the field notes (Favret-Saada and Contreras, 1981).
An attentive reading of ethnographers' field notes, particularly autho rs to hark Altern �tively, in an approach that is closer to the anthropological tradition,
example of Malinowski's notebook (1967) , has led some
that shoul d be accord ed to the process the narratIve dimensions of the ethnological study are fitted into or alternate
back to the question of the status very perso nal with analyses giving a representation of the logic of the relationships (in this
of monographic totalization (Marcus and Cushm an, 1982). The
percei ved in an almos t ritual form by case, the framework surrounding enunciations about sorcery) encountered by
nature of the act of totalization, with solitu de. the ethn�grapher (Favret�Saada, 1980). This form of narrative is interesting
anthropologists, has been revealed as an exper ience associ ated
necessary beca.use It does not unfaIrly remove the ethnographer from his/ her text,
For a long time, this experience in the field was perceived as acould then partlcularly if s/he is closely involved in the activities described therein. It
moment of immersion in a cultur e for which the anthro polog ist
ve. An analys is of field notes shows that also avoids limiting the enquiry to the trajectory of a specific person, without
become the legitimate representati ic text is more at le�st s�ggestin? why this experience is exemplary and in what way it
the switch from this experience in the field to the ethno graph
dual history p�ovides mformatIon about the type of relationship the people studied have
complicated than had been thought. The role of the actual indivi cultur e became WIth the world. Hence, this form of text transcends the alternative between
of the ethnographer in his/h er mann er of identi fying the
g on the iden­ a purely singular '1', characteristic of narrative ethnography, and the absence
more apparent, as did the influence of the actual work of writin varied encounters of the '�', which is typical of classic ethnography. The work of reflexivity is
tification of cultural types, based on much vague r and more .
the simpl e accou nts given in the published not hmIted to narratives in the first person. Through the study itself, and in
than might be imagined from chann els were a retrospective vision, the author becomes capable of describing, in the series
texts. Once this observation was ackno wledg ed, severa l
of events which s/he is reporting, in what capacity s/he was present or
explored. ial part what place s/he occupied in these events (and, notably, in the case of Jeanne
The first consists of claiming that the writing work itself is an essent to be a work �avret-Saada, the place of her work in the framework surrounding enuncia­
of ethnography. The ethnograph ic text is delibe rately consid ered
and events as a way of emph asizin g their cultur al hans). Here, we are dealing, in the field of the social sciences, with texts that
of fiction, stylizing people
17
16
ISABELLE BASZANG E R AND NICOLAS DODIER " OBSERVA TION
have the same force as 'evidence'. Often, these texts also suggest what part
of this evidence is representative of a more general condition, by using all the
possibilities of first-person narration and by identifying the role played by
the author.12
(\ / ' (,
THE
There is another way of looking at the aggregation of specific events collected
in an ethnographic study. The context of the events observed is considered
neither as a 'whole' to be discovered (integrative ethnography) nor as a
grounding point for an individual history (na:rat�ve .e�hnography), but a.s
a disparate collection of resources between WhICh mdividuais have to naVI­
gate. Unlike the cultural approach, we do not presume here that the resources
mobilized by people in their behaviour can be linked up to a coherent whole.
Unlike narrative ethnography, we leave behind the first-person account,
the aim being to generalize from the study. Th.is approach co� ld be describe.d
as combinative ethnography. It takes several dIfferent forms m ethnographIc
work. In point of fact, it is present from the very beginning of the interactionist
tradition.
Compared with the anthropological tradition, the originality of the early
Chicago School was that it did not necessarily integrate the da:a collected
around a collective whole in terms of a common culture, but m terms of
territory, of geographic space.13 The problem with which these sociologists
were concerned was based on human ecology: the interactions of human
groups with the natural environment and a given geograpl:ic�l milieu. T�eir
key concept, the unit of reference, so to �peak, was the bIOtlC co�mumty,
with its notion of territory.14 The main pomt here was to make an mventory
of a space by studying the different communities and activities of which it is
composed.
This kind of ethnography sought to identify certain cases (and notably
life histories) as examples of more general phenomena}5 but with quite a high
degree of freedom to move between different levels of generalization.16 Her.e
the ethnographer's participation in daily activities was seen as much: If
not more, as a way of collecting facts as of gaining access to the meanmg
of situations for the subjects being studied. In this respect, we are still some
distance away from the movement that was to become participant observation
in the 1950s (d. Platt, 1983) and the position of integrative ethnography.17
In the theory of action proposed by Glaser and Strauss (1967) as well � s i�
its methodological implications, the method known as 'constant companson
does not necessarily concern collective entities, but rather situations or types
of activity, classified by the sociologist and studied in their relationships
to each other, with a view to revealing their compatibility or the contrasts
between them. Individuals can switch from one line of activity to another
and the aim of the sociological study is to demonstrate the combinations -
whether harmonious or conflictual - between these multiple commitments.
18
ETHNOGRAPHY
This methodological orientation has also emerged within the
of new developments of the sociological theory of action. Here, in frame
the
work
contex
of 'sociological pragmatics', it is considered that individuals can be involvedt
in different 'regimes of action', that the arrangements that provide a frame­
work for these situations direct people towards certain forms of comm
but that tensions or combinations can emerge between these regimes ofitmentaction
,
(Boltanski and Thevenot 1991; Dodier, 1 993, 1995a). For examp the actual
work of a doctor could be studied as the articulation betwele,
'framings' of his/her patient (Baszanger, 1998b; Dodier, 1994;enSilvermdifferent
an,
1987). From this viewpoint, ethnography is no longer concerned with the
sea:ch for references shared by the actors, as in the integrative
It anns to take stock of the dynamic relationship between the realappro ach.
activit ies
of individuals within the framework of complex, normativEi references, which
are related to the situation and are not unified. Although the arrangements
fr.am�ng �he � ction are assumed to have a historical origin and a particu lar
dIstrIbutIOn m space, they are not automatically assigned to a culture. This
type of schema, which breaks with the concept of a collective consciousness
shared from the outset, assumes that individuals, and their actions, located
at the intersection of a non-harmonized plurality of references, are which are
examined in their existential commitment.
� commo� �haracteristic of these three types of ethnographic study (the
Ch�cago tradItJon, the form of interactionism inspired by Anselm Straus
SOCIOlogical pragmatics) is to distinguish between generalization and totalization.s,
The study method consists of accumulating a series of individual cases and
of analysing them as a combination between different logics of action
coexist not only in the field under consideration, but even within these that
individuals or during their encounters.
Accumulation and processing of these cases can be likened to an ethno­
graphic caseb�o�, which is gradually enriched by new examples display
forms of actIVIty and patterns of articulation. The research aims at produing new
c­
ing a combinative inventory of possible situations. The researcher has not
chosen an integrated field constituting a central point from which s/he will
reconstitute a collective whole. Instead, s/he circulates between severa
depending on which dimensions appear relevant in the analysis of eachl sites :
Although the researcher sometimes seeks a field that will allow him/hcase er to
study a regime of action or a form of activity in greater depth, s/he is not
surprised if this field proves to be more disparate than anticipated,
that forces him/her to take into consideration the way in which it is arelated
factor
to other forms of action. As a result, the material collected often appears as
rat�er vast corpus of textual data coming from very disparate sources, in termsa
of SItuations and the media used .1s
19
 ETHNOGRAPHY
ISABEl lE BASZAN GER AND NI COLAS DODIER
" OBSER VATION

FORMS OF TOTALIZATION AN D FORMS O F

nature of this global reference. It is shared by the others, and distinct from
what we ourselves are supposed to share given that we belong to other
In ethnography, several responses have been proposed toythesugge question of �o� wholes. Whether or not these other wholes are explicitly mentioned in the text
we aggregate our observations. Integrative ethno graph sts that I� IS is .unimportant it is their implicit presence that counts. Here, we are dealing
possible, thanks to monographic totalizations , to gain access to t �1e coll�ctlve WIth an otherness of belonging.
ethno graph y has qUIte radIcally Narrative ethnography is built on recording encounters between the
wholes that govern behaviour. Narrative the social scienc es - ethnographer and the study population within an individual story, estab··
challenged any pretension to totalization on the part of
critiqu es of totaliz ation in the social lishing a completely different relationship to the relevant whole: in this
it integrates itself in a vaster movement of ed accounts case, a biographical sequence. This sequence is built up gradually by the
sciences,19 and focuses on the production of highly individualiz graph ic practices field worker, who is revealed to him/herself in the field in the narrative work
gathered in the course of the study. Finall y, variou s ethno
alizat ion, which , startin g from work of the configuration of self, according to a schema defended in a more general
have sought to implement forms of gener as the revelation sense by Ricoeur (1990). The other individuals are perceived in the same way:
on series of cases, aim not so much at the totaliz ation of data
dispa rate or even contra dictor y forms as so many individual beings also involved in the process of' the construction
of a combinative mechanism between of self, caught up in the same tangled web of stories in the field.
of action within a given society. back to a non- Closure occurs when the ethnographer deems that s/he is capable, at a
Methodologically speaking, these last-mentioned works refer in the Chicago given moment, of giving his /her experience the shaping form of the narrative.
cultural concept of the field which was alread y presen t
r conce rn for theore tical clarification Closure does not rule out further developments that might spring from
tradition. However, here there is a greate and processing later reflection by the fieldworker about him/herself, or further applications
concerning the concept of action and the metho ds for gather ing
data.20 This avoids both the myste ries of mono graph ic totaliz ation and the which might be directed to him/her by contacts in the field. Here the reader
n of the fieldw orker eviden t in some narrat ive is faced with the otherness of the ethnographer as much as that of his/her
excesses of focusing on the perso ivity. contacts. At the centre of this otherness we find the interplay between two
ethnographic works, or the escap e back towar ds hyper -reflex
What is the aim of each of these forms of ethnography? In thishow chapter, we forms of identity proposed by Ricoeur, 'ipseity' and sameness. Ipseity comes
will look at this question from the persp ective of othern ess: do they f�om the Latin pronoun ipse and defines a form of identity referring to the
ity? We sugge st that these forn : �thnog­
s of smgular person, which is distinct from 'sameness', a form of identity referring
represent the fact of human plural
raphy propose what we might call differe nt mode s of othern e � I..e. dIfferent
s, to categories of belongingness of the person. The interplay between these
the peopl e presen t and actmg m the texts two forms of identity as a characteristic of the narrative leads, as he suggests,
ways of showing the reader how to the narrative identity.
are both similar to and different from him/ her.
The relationship of individuals to the wholes or aggregatesare of integra­ Integrative ethnography proposed that the ethnographer disappear to
all tools reveal better to the reader an otherness of belonging (that of the natives).
tive ethnography is one of belongingness. These aggregates'belong', so to Conversely, in narrative ethnography, the presence of the ethnographer
for distinguishing among indivi duals betwe en those that
fication mark�rs, confronts the reader with a narrative otherness, grounded in the 'ipseity'
speak, and those that do not. Hence, they are social identigates of claSSIcal of the ethnographer but none the less - thanks to the very dynamic of the
within the meaning given by Goffm an (1961) . The aggre
sedly narrative offering possible avenues towards generalization beyond the
ethnography, as noted by Piette (1996), correspond to objects suppo individual case.
'shared' by people. These are shared only by the people contra presented in the
ethnographic text, and not by its readers, who are, on the there' ry, invited The question of belonging, which is central to integrative ethnography,
to see in ethnography faithful images of what happens 'overonly recogn , and not and that of individual biographies, central to narrative ethnography, is
'here' (Geertz, 1988). Moreover, these shared objects are erself doesized not
replaced in combinative ethnography by the question of action. The initial
point of entry into the field is generally an enquiry into a form of activity,
as such by the ethnographer because the ethnographer him/h ex�l�red in different places: disputes, judgement, scientific activity, medical
share them. closure act!vI.ty, production, etc. In this case, the ethnographic text is very similar
The whole process of assembling observations, up to and including only by a to a kmd of inventory. It is generally presented as a list of possible operations:
of the study, is based on the fact that what is shared is shared ethnographer the resources available to people to act while taking into account situational
circumscribed group of individuals which includes neither the one identifies constraints.
nor, in most cases, the people who read ethnography. Whether(Levi- Strauss), Nevertheless, despite the basically disparate or even conflicting nature
the recording of this aggregate in an integrative experience in a successful of the forms of action envisaged, these inventories do not contradict the idea
in a hermeneutics of indigenous interpretations (Geertz) orreflect s the dual that the people involved act with reference to 'shared objects',21 However,
socialization (Fox), the very fact that closure can be reached
21
20
ISABELLE BASZANG ER AND N I COLAS DODIER OBSERVATION
these objects do not have the same status as in integrative ethnography. �he
resources mobilized are presented very specifically in the form of a relatIve
availability. They are available to people because they belong to a common
fund of skills, shared by all those appearing in the text.
This availability is relative since mobilization of these resources impl�es
a cost, unless they become available owing to some unexpected opportumty
or development. These resources, which are potentially available to indi­
viduals, do not, as in integrative ethnography, 'absorb' these individuals, to
use the concept developed by Piette. They are too disparate or even mutually
contradictory for that.
In combinative ethnography, the coherence of resources is lost, but not the
existence of a shared aggregate of dispositions for understanding the world.
This aggregate is presented as a common fund of possible actual�zable
operations. The origin and contours of this common fund vary accordmg to
the forms of activity studied and the underlying frameworks.
So we find studies referring to a common fund shared by all (or virtually
all) human beings. These studies, while showing disparate forms of ac.tion,
none the less relate them to a global reference as a source of shared objects.
But it is a very general reference, not particularized as in the case of integrative
ethnography.
Ethnomethodology refers to the skills of 'members' as shown in their mastery
of natural language. Through this mastery, people are capable of dem�n­
strating, in situ, the accountability of their activities. The work of Sebastlan
McEvoy on 'defensive invention' (1995) attributes this very general skill
in handling language in a given situation to people involved in sequences
of accusation and defence and in the myriad possible ramifications of the
language in which disputes may be expressed. Luc Boltansk� a�d Laurent
Thevenot refer explicitly to the existence of a very general skIll m peopl�'s
commitment to multiple forms of action directed towards a concern With
justice, renewing both for philosophy and history, the question of the genesiS
of these skills.
In other studies, the status of this common fund is more flexible. Erving
Goffman, for instance, refers to interactional skills which are shared by all
individuals in their daily lives, and skills available to people under more
specific conditions. One of the attractions of his work lies in the connections
he creates between very different places, even in work which, in other respects,
as in Asylums (Goffman, 1961), presents a monographic-type unity of time
and place. The contours of the common fund of resources tend to be more
strictly demarcated in the interactionist tradition inherited from Anselm
Strauss, notably in the sense that concepts such as 'social world' (Strauss, 1978)
tend to refer to circumscribed groups of people brought together around
a shared horizon of action.
Hence it is possible to build up a theory that occupies an intermediate
position between combinative and integrative ethnography, oscillating
between the study of a common fund of interactional skills in our societies
and a demonstration of the existence of varied funds of knowledge distributed
22
ETH NOGRAPHY
between different groups of people. Here, at the extreme, skills are so plastic
that they depend completely on the eminently local context surrounding the
person, social distinctions between individuals on the basis of their belong­
ingness once again becoming meaningless. This is shown in the follow-up
work on scientific innovation carried out by Michel CalIon and Bruno Latour:22
their narratives propose a vision of a world in which constant movement in
the chain of socio-technical networks is echoed by transformation of the skills
of the human beings engaged in these networks. The skills referred to are, once
again, common, but in the very general sense of a capacity to maintain and
keep in operation the networks of sciences and techniques.
This dual language of 'common' skills and 'possible' acts reflects the new
status ascribed to closure of the ethnographic field. Here, inventories and
casebooks are associated with the principle of a closure that is always
envisaged as relative. In Glaser and Strauss's 'grounded the�ry', closure corre­
sponds to a saturation of data within the framework of gradual sampling
operations that take into account unexpected developments in a collection
organized in different places, which are then compared with each other
(Glaser and Strauss, 1967) .
In the work of Boltanski and Thevenot (1991), the forms of possible actions
are, to some extent, closed off by the enumeration of 'cities' to which actors
are assumed to make reference when addressing the concept of justice, but
the list of figures inside the model increases as observations in the field feed
the 'common worlds' concretely linked to the cities. Moreover, the model does
not rule out the historical emergence of new cities supplementing those that
already exist, and interfering with them in unexpected combinations.23 Francis
Chateauraynaud's work on professional mistakes (1991) and, in association
with Christian Bessy, on the expertise of objects (Bessy and Chateauraynaud,
1995), is characterized by an epistemology that resolutely emphasizes the
endless deployment of figures that can be demonstrated from interrogation
into a given form of activity, consistent with the inventions that people may
propose.
The relativity of closure of an inventory of possible actions is exacerbated,
in the ethnography of regimes of action, by the fact that each observed 'case'
is itself considered to be open. Once we accept that each protagonist has
possible resources to attempt to change the temporal outcome of the action,
there is nothing (apart from the constraints specific to empirical enquiry itself)
to justify closing off at any particular point following the dynamics of stories
that might continue to be played out in the field and elsewhere, even in the
most unexpected configurations.
Here we can see where combinative ethnography differs from the preceding
forms of ethnography in terms of the mode of otherness proposed to readers.
Otherness is not seen in the same way as in integrative ethnography, i.e. in terms
of belongingness. No distinction is made between people as such; that is, in
terms of their acquisition of dispositions differentiating them in a sufficiently
stable manner inside shared, albeit circumscribed, wholes. In the case of those
ethnographic works which take the most radical line, we are, as readers, authors
23
 ETHNOGRAPHY
ISABElLE BASZAN GER AND N I COLAS DODIER § OBSERVATION

and the people encountered in the field, assumed to share a common fund of Ultimately, combinative ethnography presents itself as a vast inventory
skills, and it is this fund revealed by the analys is which is deploy ed along­ of possibilities or potentialities regarding situated actions. Each new study
side the other entities present in the world: artefac ts, natura l entitie s. Or, as is observes new scenes and hence helps enlarge the spectrum of skills, arrange­
the case in work which allows a wider margi n for distin guishi ng betwe en ments and forms of action explored by earlier studies. The discourse of
people in terms of a flexible belongingness, the cont ?� rs of comm on funds and possibilities characteristic of these works springs from taking into account two
their inscription within groups are not clearly specIfIed. simultaneous properties of action: individuals have to deal with situational
. Imply
. constraints; but they always have the possibility of dealing with them by
This reference to a common fund of resources does not necessarIly e, but that redefining the situation. However, these possibilities impose a cost for the
that we cannot differentiate between the positions taken by peoplare defined people using them. Such costs can involve internal tensions, linked to the
they are first and foremost action related . This means that they
. of theIr. encounter
by the respective positions occupied by people at the tlme individual's confrontation with heterogeneous forms of activity, with HIe work
rather than by their belongingness to a given group or aggreg ate: such of adjustment required to move from one form to the other, or to harmonize
as victirn s/benefactors/ spectators; care worke rs / patien ts; judges / prison ers/ one with the other in the course of the action. All the cases show the efforts
counsel for the defence, and so on. made by the individual to reorganize around him/hers�lf the other people
Uneven distribution of skills between people is envisaged in terms of these and arrangements engaged in the same situation.25
respective positions, rather than as a functi on of what each party has acquir :d This brings us to the question of creating an equivalence between these
through belonging to distinct aggreg ates. Here, ethnog raphy seeks to buIld costs: how do we measure possibilities and establish a comparison between
a sociology of encou nters24 between persons occupying different action -related them'? Crucial to the different forms of combinative ethnography is the
positions with respect to an 'anthropolog ica ! ' proble m (i.e. linked to a
. of a fund of SkIlls
gene.ral representation of fundamental possibilities or potentialities. Even if they
human condition). People are seen to be 111 posses slOn have not hitherto been actualized, the skills referred to here are considered
offering various possibilities of commitment in the encounter, dwhile rem�i�i�g as being potentially in people, who draw on them primarily as a function
within the constraints fixed by the arrang ement s create by the 111Itlal of situational opportunities. The paramount factor in whether they deploy
situational context. these potentialities is whether they have or have not yet encountered the
arrangements or action-related positions allowing them to do so. The charac­
teristics of people who have acquired distinct sets of capacities through past
WHAT IS
experience is not systematically investigated.
This does not mean that combinative ethnography is not interested in the
This suggests the emergence of a new mode of otherness which we might question of acquisition of skills. Indeed, a number of researchers have reflected
call pragmatic otherness. It proceeds from the fact that the reader is brought on the comparison between distinct processes of acquisition, linked to distinct
into contact, through the text, with individuals who share with the reader forms of action.26 However, in these works, differentiation between persons
and between themselves a common fund of heterogeneous resources and in terms of what they have actually mobilized is of secondary importance.
who may, depending on the case, occupy varied action-related positions. This assessment of potentialities tends to compare the different situations with
From the reader's viewpoint, and despite all the myriad ethnographic which a generic person might be confronted.
details provided, the text does not present people that are 'differ�nt', either This manner of looking at individuals reflects a theoretical choice: the
in terms of their acquired skills as in integrative ethnography, or 111 terms of primary aim is to construct a theory of potential skills, actualized depending
their individual stories, as in narrative ethnography. Indeed, the criticism on the encounters of each individual with other people, with arrangements
of 'disembodiedness' frequently levelled at the most radical texts of combi­ and with action-related positions, rather than a theory of differentiated
native ethnography reflects their failure to satisfy this implicit expectation of acquisition of skills. However, it also highlights the limitations of combinative
integration of observations in persons distinguished from one another. ethnography as a study method. Integrative ethnography proposed a simple
Whereas successful integrative ethnography offers its readers an encounter and attractive schema, marrying close observation of behaviour with a
with appreciably different individuals, combinative ethnography offers recording of the differences between people. Measuring possible actions
an explicitation of what is in fact present in all of us, albeit not nec� ssarily was comparatively clear: the society or the unit of belongingness defined for
activated owing to the limited opportunities afforded us by our sItuated each person the range of appropriate possibilities, those of their society or
commitments. It pictures a world of common or indeed plastic capacities, with group, and excluded other possibilities as so many reactions incompatible
the potential for indefinite modelling in reaction to the sudden appearance with this belongingness.
of the non-human objects with which we are confronted owing to the incessant Looked at from the perspective of heterogeneous worlds, the skills acquired
transformations of socio-technical networks. by people emerge less clearly. Here, we need to 'follow' individuals, and

25
24
ISABEllE BASZANGER AND N ICOLAS DODIER " OBSERVATION
observe the very complex way in which the effects of successive con:mitme�ts
in composite situations are embedded in these people. There IS nothmg
to prevent an ethnographer undertaking this follow-up of individuals and
indeed this is the process found in case histories centring on a given person,
whose trajectory is recounted in the light of salient. moments .abserve� �y' t�1e
ethnographer.27 Nevertheless, a genuine comparatIve analysIs of possibilItles
linked to the differentiated positions acquired by individuals presupposes the
establishment of a common space within which the facts relating to their
behaviour would be systematically recorded.
This work is not incompatible with ethnography, but it does require some
additional investment at a given moment in tools capable of representing
in a single item (a diagram of connections, a statistical t�ble, etc.) �h� often
quite large numbers of people concerned in terms of the kl11ds of actlvity that
constituted the point of entry into the study. Only in this way can we hope
to visualize the degree to which the positions occupied by individuals are
similar or different and compare the possibilities available to individuals
structured in a distinct manner within a single space of coexistence?8
Within such a space, individuals emerge according to a new form of other­
ness: the otherness of acquired skills. The reader is introduced to people
who over time have acquired distinct sets of potentialities. This form of other­
ness provides a key for reflecting on acquired skills in heterogeneous worlds,
beyond the schema of belongingness proposed, for other contexts, by
integrative ethnography. . . .
. the public space the elements constituting the (often hidden) pragmatic
Representation of a common space of coexlStence necessanly ImplIes
selecting a few fairly simple traits characteristic of the actions of the study
population, given the complexity of the operations revealed by ethnog­
raphy of action. The degree to which this activity is simplified depends on
the resources which the fieldworker is able to invest and the limitations,
at any given time, of the processing tools available t� deal wi�h �omplex
bodies of data.29 In addition, the different forms of commItment comcide more
or less exactly with the representational arrangements required to visual.ize
these common spaces. Certain forms of action can hardly be recorded (DodIer,
1990) and other forms can only be recorded in very specific forms of writing
as Boltanski showed in his study of love as agape (1990).
Here we can measure the very reductionist effect on the pragmatic condition
of individuals of ethnographic observations aimed at illustrating work which
tries to represent systematically the differentiated positions acquired by
individuals in a given society - notably in the sociology of Pierre Bourdieu.
Indeed, in these observations, the concept of pragmatic otherness disappears,
practically speaking, when all the skills of commitment in situations are
related, as a matter of principle, to so-called social positions.
Ethnographic observations then exist, first and foremost, as a means of
exemplifying phenomena objectified elsewhere, using other methods, mainly
statistical ones. The lucidity required to unmask in each situation what
is related to social positions monopolizes the interpretive effort. There is no
longer any possibility to reflect on what in the relationship with the other
26
ETHNOGRAPHY
,
is related to a genuine pragmatic otherness and is not simply the projection,
at the moment of the encounter, of positions acquired in the specific form of
the common space of coexistence that is the social space engendered by the
forms of capital uncovered by Bourdieu. The question of differential
assessment of possibilities for acting by virtue of acquired positions is assumed
to be already resolved even before the relevant forms of action are brought
to light in the study.
ETH N O G RAPHY IN T H E P U B LIC SPACE OF ITS R EADERS
One contribution made by combinative ethnography is its capacity to explain
to readers the skills for acting with which they themselvj?s are unfamiliar as
objects of reflection. Every person can see him/herself as potentially concerned
both by situational constraints and by possibilities for acting. Moreover,
the close observation of activities which ethnography allows here possesses
a capacity to bring to light, from a critical point of view and in a public space,
skills that are not acknowledged by the usual arrangements for recording
activities. In particular, close observation of individual behaviour uncovers
the texture of the activity by revealing the multiple operations that individuals
perform in order to act in a complex universe.
Indeed, one aim of this work of combinative ethnography is to bring into
condition of individuals. This type of ethnography renders visible to an audi­
ence a whole complex web of activity which otherwise would be apparent
only to those engaged in the situation, at the time and in that place, generally
without leaving any durable and transportable trace of their commitment.3D
Integrative ethnography brought into close perspective the fabric of the
life of individuals belonging to other societies. Combinative ethnography
shows us what we all mobilize in the course of action, or what we might
be brought to mobilize if confronted with a given set of arrangements or a
given action-related position. Its real aim is to bring to light what generally
remains invisible in official accounts, both by revealing the work of adjustment
and by acknowledging skills for acting that are typically underestimated.
The critical compass of ethnography is very sensitive, for instance, to
what ethnomethodology and laboratory-based ethnographies reveal concern­
ing the local adjustments of scientific practice which escape the remote
observation formerly allotted to it by epistemology. What is true of skills is
als? true of the pragmatic tensions experienced by people in a heterogeneous
umverse. These tensions are only revealed by close observation of behaviour
which force us to look beneath the smooth surface that some actors ascribe
to their organizations when they undervalue, consciously or not, their costs
for concrete activities.
Hence this work involves explaining concrete conditions for acting,
op��ness to new capacities and a critique of what is ordinarily hidden in
offICial accounts. This critique proceeds not so much by elucidating the real
27
ISABELLE BASZANG ER AND NICOLAS DomER {, OBSERVATION
interests of the actors, as in much sociological work, but by revealing the
adjustments that people are obliged to make, whatever their other interests,
in the detailed course of their actions. Its characteristic - and also its limitation
is that it concerns itself with people who are not differentiated from each
other in terms of acquired skills and of whom we can, in the last analysis, say
that they are not collectively structured in terms of their differences.
This is the point at which, in an extension of research into pragmatic
otherness, we might usefully undertake work concerned more with the
otherness of acquired skills, which would seek to represent distinct positions
in a common space of coexistence. Making reference to such a space should
give readers the possibility, which combinative ethnography has temporarily
neglected, of situating themselves with respect to others, as persons struc··
tured differently by time and not only as persons endowed with similar
potentialities. Representing such a space also answers the concern to totalize
ethnographic data in the framework of a collective entity, once the schemas
of belongingness typical of integrative ethnology have been abandoned.
N OTES
1. 'To treat facts of a certain order as things is not then to place them in a certain
category of reality but to assume a certain mental attitude toward them on the
principle that when approaching their study we are absolutely ignorant of their
nature, and that their characteristic properties, like the unknown causes on which
they depend, cannot be discovered by even the most careful introspection'
(Durkheim, 1982).
2. This last argument is reinforced, on the epistemological leve!, by the fact that the
negotiated character of the design and implementation of an experiment has, in
any case, generally come to light thanks to work in sociology of science. Even
in the most detailed experimental plans, negotiations are often eliminated from
reports, in spite of the fact that they actually exist (Collins, 1985; Knorr-Cetina, 1981;
Latour and Woolgar, 1979).
3. This is why themes of duplicity, treachery and manipulation are at the heart of
such narratives (Leiris, 1981). An implicit part of the ethnographer's condition is
that slhe has to resolve these tensions as they appear.
4. On the opposition between monograph and statistics, see Desrosieres (1998).
5. 'There is a general tendency to think of a study based on participant observation
as largely the product of an esoteric, personal kind of clinical talent on the part of
the fieldworker, who is considered to be endowed with qualities usually referred
to as "sensitivity", "intuition" and "empathy'!! (Fox, 1974: 231).
6. This separation is achieved via an initial work on oneself. 'On several afternoons
and evenings at Harvard, I found myself considering a trip to Cornerville and
then rationalizing my way out of it. . . . Then too, I had to admit that I felt more
comfortable among these familiar surroundings than I did wandering around
Cornerville and spending time with people in whose presence I felt distinctly
uncomfortable at first. When I found myself rationalizing in this way, I realized
that I would have to make a break. Only if I lived in Cornerville would I ever be
able to understand it and be accepted by it' (Whyte, 1981: 293-4).
7. It is clearly shown in the example in the appendix of Bosk (1979).
28
The way in which Renee Fox describes her understanding (discovery) of the
8.
meaning and function of black humour for sick people nicely illustrates this
phenomenon: 'At a non-hospital gathering one evening, I caught myself in the act
of �aking a macabre joke, and I can remember speculating on the source of my
unlIkely new talent. The next morning, as I moved to [the ward], for the first time
I n oti�ed how much of the ward's conversation was phrased in the language of the
gnm.
Joke and how often I responded in kind . Without realizing it, I had learned
to speak to the men of [the ward] in the same way that they talked to each other.
Long before this insight occurred, my field notes contained many samples of ward
hu�nor. But it was only by virtue of self-observation that I became suffiCiently aware
of Its prevalence to regard it as a phenomenon central to my study' (1974: 231;
emphasis added).
9. !his point appears clearly in the relationship between the initial field notes, which,
m many ways, already contain 'everything', and the final,analysis which came
months later: 'From the very start . . . my notes contained almost all the components
of the ward picture I was ultimately to assemble. However, at the time that I
recorded this observation, I was not yet aware of the patterned interconnections
between them. At what point did I begin to see the ward in a coherently structured
way? In the sense of month and day, I cannot really answer that question. But I do
know tha � the so-called "understanding" of [the ward] which I eventually attained
was not SImply the result of coming to know more about the ward in a cognitive
sense. It also involved a process of attitude learning (very much akin to what social
,
scientists mean when they refer to the process of "socialization") (Fox, 1974:
217-18).
10. We can, for example, refer to the classical manuals of Griaule (1957), Maget (1962)
or Mauss (1947).
11. Se� also Certeau (1987) concerning the relationship between history and narrative.
12. EVIdence about working-class lifestyles are remarkable in this respect in their
capacity to combine accounts of individual histories anchored in very specific
contexts and preservation of a general framework of discussion (Linhart, 1978;
Navel, 1945; Wei!, 1951).
13. The sociological objective of this tradition was an attempt to understand the new
urban space emerging as a result of industrialization and the double phenomenon
.
of ImmIgratIon that accompanied it (from the south and rural zones towards the
nort� and t�e dawning metropolises, on the one hand, and from the European
contment wIth Its multiple ethnic components, on the other). Chicago, with its
mosaic of ethnic groups and its different socio-ethnic neighbourhoods, was
emblematic of this phenomenon. The aim was to analyse the ways in which this
space in which different groups confronted each other and mixed together was
structured and to study their reactions to these totally new living conditions.
1 4. For �xample, Thrasher (1927) started from the question of the geographical
localIzation of juvenile delinquency: some sectors are more affected than others;
how can this be explained?
15. The preferred approach is the s�udy of natural history followed by a study of
.
the cOl�mulUty (understood here 111 the sense of a biotic community with its notion
of terntory): a town, a neighbourhood, an ethnic community located in a given
.
geographIcal space (the ghettos). The central method used is the case study, which
IS based essentially on a life history and, to a lesser extent, on non-structured
mterviews using the actual words of the subject and all sorts of personal documents
(personal letters, evidence collected in community notebooks, etc.).
29
ISABelLE BASZANGER AND N I COLAS DODIER OBSERVA TION
16. The important point was to gather intensive data for example, for The Hobo,
Anderson (1923) collected sixty life histories, and made a preliminary study of
400 tramps, one use of which was to establish a list of the apparent physical defects
of hobos and to identify both individual and more general traits. Alongside these
life histories, the facts were collected via a study of administrative statistics,
archives, local newspapers and the case files of social workers.
1 7. The fieldworker's position is different from the empathetic position of the
participant-observer and is closer to that of the stranger, if we refer to Sombart and
especially to Simmel. In a text written during his 1927 survey and published in
1983, Cressey shows how he attempted to build up a position as stranger in his
ethnographic relationships to study the environment, the world of taxi dance halls.
Referring to Sombart's analysis of 'the cultural stranger', he distinguishes between
two 'stranger positions' which he used in his research. The 'sociological stranger'
is a stranger with a particular status as commonly used by lawyers, doctors, social
workers, public school counsellors, etc. The second, which he used much more
often, is that of the 'anonymous stranger'. This is an ordinary relationship in
big cities where isolated people meet up in transient relationships and with free
time on their hands. For the researcher: 'it provides an opportunity for exploring
aspects of human nature not ordinarily revealed' and has the effect of a 'catharsis'.
This ethnographic relationship allows the field worker to access revelations that
Cressey calls 'impersonal confessions'.
18. The work of processing these data can benefit from the development of automated
techniques designed to facilitate constant two-way communication between the
encoding of the material and ethnographic concentration on its special features
(Chateauraynaud, 2003).
19. On the critique of statistical totalization and its limitations, see Dodier (1996).
20. See, for example, the enumerations made in Becker et al. (1961), and, in a more
general sense, for the relationships between qualitative and quantitative data in
theoretical elaboration, see Glaser and Strauss (1967: Chap. 8).
21 . For developments on this point, and more particularly concerning interactionism
and ethnomethodology, and taking a critical approach to this referring back to
shared objects, see Piette (1996).
22. See for instance Calion (1998) and Latour (1987).
23. It is worth pointing out that in a recent work Boltanski and Chiapello (1999) analyse
the emergence of a new city 'the city of projects' (la Cite 'par projets'), linked to the
latest developments of capitalism.
24. The encounter is characterised by the fact that the people involved are in each
other's horizon of action, whether directly or indirectly through objects of com­
munication. At the very least, these people are in each other's horizon without
this necessarily being reciprocal, as in the scene analysed by Boltanski (1999)
based on an action-related schema, of the television viewer faced with the sight of
another person's suffering. Encounter-driven relations can be distinguished from
relations of interdependence, which includes all those situations where the action
of one person has consequences for another, without one being necessarily present
in the horizon of the other (see Corcuff, 1995).
25. See for instance the work of Goffman (1974) on shifts in 'framing', the analyses of
Strauss et al. (1985: 151-90) on the 'work of articulation', or the revelation of the
cost of moving between different 'common worlds' in Boltanski and Thevenot's
model (1991).
26. We might note the observations of Thevenot (1994) on the constitution of familiarity
30
P
ETHNOGRA HY
in the relationship to objects, and the way in which Bessy and Chateauraynaud
(1 �95) use the concept of 'hold' to describe the embedding, in both persons and
thmgs, of moments when the body is engaged at a level that does not mobilize
exercise of judgement.
27. A good example is Strauss et aJ.'s development (1 985: 161--81) of the case of
Mrs Price in the framework of an analysis of the trajectories of cumulative mess in
the hospital environment.
28. This space may bring together contemporaries who live at the same point in history.
�t n:a,Y also aim to grasp the effects of history on the internal structuring of
mdividuals, notably to situate observations relative to the present in an h.istorical
perspective (Baszanger, 1998b; Young, 1995).
29. These limitations are fortunately tending to recede thanks to the recent develop­
ment of automated techniques allowing a continuous movement back and forth
between encoding of material and an ethnographic attentiop to its particularities
(Chateauraynaud, 2003).
30. We might think here of the way in which Anselm Strauss's focus on the different
d�mensio�s of medical work and above all the difficulty of harmonizing the
dIfferent lmes of work allowed him to see and bring to light the work done by
the patient, in all its different dimensions (Baszanger, 1998a). For an argument
supporting the need for close observation in order to apprehend the intensity of
the web of judge�ents of ability between operators responsible for ensuring the
.
operation
of techmcal systems, see Dodier 0995b).
Recommended r..<,rii ... "
REFER ENCES
Anderson, N. (1923) The Hobo. Chicago: University of Chicago Press.
Arendt, H. (995) Qu'est-ce que la politique? Paris: Seuil.
Baszanger, 1. (1998a) 'The work sites of an American interactionist: Anselm Strauss,
1917-1996', Symbolic Interaction, 21 (4): 353-77.
Baszanger, 1. (1998b) Inventing pain medicine. From the laboratory to the clinic. New
. NJ: Rutgers University Press.
BrunswIck,
Becker, H., Geer, B., Hughes, E. and Strauss, A. (1961) Boys in white. Students' Culture
In MedIcal School. Chicago: University of Chicago Press.
31

ISABEllE BASZANGER AND N ICOLAS DODIER it OBSERVATION
Bessy, C. and Chateauraynaud, F. (1995) Experts et faussaires. Pour une sociologie de la
perception. Paris: Metailie.
Boltanski, L. (1990) L'Amour et la Justice comme competences, Trois essais de sociologie de
l'action. Paris: Metailie.
Boltanski, L. (1999) Distant suffering: morality, media, and politics. Cambridge and New
York: Cambridge University Press.
Boltanski, L. and Chiapello, E. (1999) Le nouvel esprit du capitalisme. Paris: Gallimard.
Boltanski, L. and Thevenot, L. (1991) De la justification. Les economies de la grandeur. Paris:
Gallimard.
Bosk, C. (1979) Forgive and remember. Managing medical failure. Chicago and London:
University of Chicago Press.
Calion, M. (ed.) (1998) The laws of the markets. Oxford and Malden, MA: Blackwell/
Sociological Review.
Certeau, M. de (1987) Histoire et psychanalyse entre science et fiction. Paris: Gallimard.
pp. 66-96.
Chateauraynaud, F. (1991) La faute professionnelle. Une sociologie des conflits de
responsabilite. Paris: Metailie.
Chateauraynaud, F. (2003) Prospero. Une technologie litteraire pour les sciences sociales .
Paris: Editions du CNRS.
Clifford, J. (1 983) 'De l'autorite en tehnographie', L'Ethnographie, (2): 87···118.
Clifford, J. and Marcus, G. (1986) Writing culture: the poetics and politics of ethnography:
a School of American Research Advanced Seminar, Berkeley, CA: University of
California Press.
Collins, H. (1985) Changing order: replication and induction in scientific practice. London:
Sage.
Corcuff, P. (1995) Nouvelles sociologies. La realitC sociale en constructions. Paris:
Nathan.
Cressey, P. (1983) 'A comparison of the roles of the "sociological stranger" and the
"anonymous stranger" in field research', Urban Life, 1 2, (1): 102-20.
Darbo-Peschanski, C. (1987) Le discours du particulier. Essai sur l'enquete hCrodoteenne.
Paris: Seuil.
Desrosieres, A (1998) The politics of large numbers: a history of statistical reasoning.
Cambridge, MA: Harvard University Press.
Dodier, N. (1990) 'Representer ses actions. Le cas des inspecteurs et des medecins du
travail', Raisons Pratiques, (1), Les formes de l'action, 1 15-48.
Dodier, N. (1 993) 'Acting as a combination of common worlds', The Sociological Review,
41 (3): 556-71.
Dodier, N. (1994) 'Dealing with complaints in expert medical decisions. A sociological
analysis of judgment', Sociology of Health and Illness, 16 (4): 489-514.
Dodier, N. (1995a) 'The conventional foundations of action. Elements of a sociological
pragmatics', Reseaux. The French Journal of Communication, 3 (2): 147-66.
Dodier, N. (1995b) Les hommes et les machines. La conscience collective dans les societes
technicisees. Paris: Metailie.
Dodier, N. (1996) 'Les sciences sociales face a la raison statistique', Annales. Histoire et
Sciences Sociales, (2): 409-28.
Durkheim, E. (1982) The rules of sociological method. New York: Free Press.
Dwyer, K. (1979) 'The dialogic of ethnography', Dialectical anthropology, 4 (3):
205-24.
Favret-Saada, J. (1980) Deadly words: witchcraft in the bocage. Cambridge and New York:
Cambridge University Press.
32
r
ETH NOGRAPHY
Favret-Saada, J. and Contreras, J. (1981) Corps pour corps, Enquete sur la sorcellerie dans
Ie bocage. Paris: Gallimard.
Fox, KC. (1974) Experiment Perilous. Philadelphia: University of Pennsylvania Press
(first published in 1959).
Gadamer, H.G. (1975) Truth and method. London: Sheed & Ward.
Geertz, C. (1 973) The interpretation of cultures. New York: Basic Books.
Geertz, C. (1988) Works and liDes: the anthropologist as author. Stanford, CA: Stanford
University Press.
Glaser, B. and Strauss, A (1967) The discovery of grounded theory. Chicago: Aldine.
Goffman, E. (1961) Asylums: Essays on the situation of mental patients and other inmates.
Garden City, NY: Anchor.
Goffman, E. (1974) Frame analysis: an essay on the organization of experience. New York:
Harper & Row.
Griaule, M. ( 957) Methode de l'ethnographie. Paris: PUF. ,
Knorr-Cetina, K. (1981) The manufacture of knowledge. Oxford: Pergamon.
Latour, B. (1987) Science in action: how to follow scientists and engineers through society.
Cambridge, MA: Harvard University Press.
Latour, B. and Woolgar, S. (1979) Laboratory Life. The production ofscientific facts. London:
Sage.
Leiris, M. (1981) L'Afrique fan tome. Paris: Gallimard (original edition 1934).
Levi-Strauss, c. (1963) Structural anthropology. New York: Basic Books.
Licoppe, C. (1996) La formation de la pratique scientifique. Paris: La Decouverte.
Linhart, K (1978) [,'etabli. Paris: Minuit.
McEvoy, S. (1995) L'invention defensive. Paris: Metailie.
Maget, M. (1 962) Guide d' etude directe des comportements culturels (ethnographie
metropolitaine). Paris: Editions du CNRS.
Malinowski, B. (1967) A diary in the strict sense of the term. London: Routledge & Kegan
Paul.
Marcus, G . and Cushman, D. (1982) 'Ethnographies as texts', Annual Review of
Anthropology, 1 1 : 25-69.
Mauss, M. (1947) Manuel d' ethnographie. Paris: Payot.
Moerman, M . (1972) 'Analysis of Lue conversation: providing accounts, finding
breaches, and taking sides', in D. Sudnow, (ed.), Studies in social interaction. New
York: The Free Press. pp. 1 70-228.
Navel, G. (1945) Travaux. Paris: Stock (republished by Folio Gallimard).
Piette, A (1996) Ethnographie de I'action. L' observation des details. Paris: Metailie.
Platt, J. (1983) 'The development of the "participant observation" method in sociology:
origin, myth and history', Journal of the History of the Behavioral Sciences, 19: October.
Ricoeur, P. (1984-1988) Time and narrative. Chicago: University of Chicago Press.
Ricoeur, P. (1990) Soi-meme comme un autre. Paris: Seuil.
Ricoeur, P. (1991) From text to action. Evanston, 11: Northwestern University Press.
SIlverman, D. (1987) Communication and medical practice. Social relations in the clinic.
London: Sage.
Strau ss, A (1978) 'A social world perspective', in N.K. Denzin (ed.), Studies in symbolic
II1teraction . Greenwich, CT: JAr Press. Vol. 119-28.
Strauss: A, Fagerhaugh, S., Suczeck, B. and Wiener, C. (1985) The social organization of
medIcal work. Chicago: University of Chicago Press.
Thevenot, L. (1994) 'Le regime de familiarite. Des choses en personne', Geneses, 17:
72-1 01.
Thrasher, F.M. (1927) The Gang. Chicago: University of Chicago Press.
33
ISABelL E BASZAN GER AND NICOLAS DODIER OBSERVATION

Weil S. (1951) La condition ouvriere. Paris: Gallimard.

;
. .
Umversity of
Wh te, W.F. (1981 ) Street corner society, Third Edition. Chicago: The
Chicago Press. . .
. stress dIsorder . Building bridges
Young, A. (1995) The harmony of illusions. Inventmg post-traumatIc
Princeton, NJ: Princeton University Press.
The possibility of analytic dialog ue
between ethnog raphy, conversation
analysis and Foucault

Gale Miller and Kathryn J. Fox

This chapter extends and elaborates the analytic potential of qualitative

research by considering how it may be used to construct bridges between
different approaches to social life, particularly perspectives that focus on
macro- and microscopic issues. The analysis deals with the ways in which
the microsociological insights of ethnomethodology (Garfinkel, 1967; Heritage,
1984; Mehan and Wood, 1975; Zimmerman, 1969) and conversation analysis
(Atkinson and Heritage, 1984; Boden and Zimmerman, 1991; Button and Lee,
1987; Sacks, 1992; Sacks et al., 1974) may be linked with the macro-historical
emphasis of Foucauldian discourse studies (Dreyfus and Rabinow, 1982;
Foucault, 1972, 1980; Lindstrom, 1990; Shumway, 1989). Our aim is to explicate
the parallels between the various approaches, bringing both their differences
and similarities into relief. Rather than promising an exhaustive discussion
of each methodological approach, this chapter will explore particular relevant
features of each approach that inform the associations we make. It is our hope
to show how aspects of Foucauldian scholarship, for example, share common
themes with ethnomethodology and conversation analysis.
We use the bridging metaphor self-consciously. Bridges link distinctive
land formations, making it possible for people to traverse between them. While
opening new opportunities for residents on each side, bridges do not blend
the formations or otherwise make them indistinguishable. Where possible,
bridges are also built to span the shortest distance between the land forma­
tions. The same conditions hold for this analysis. Our purpose is to show how
two or more analytic formations may be linked and made mutually infor­
mative, while also respecting the distinctive contributions and integrity of
each perspective. The analysis is also intended to identify the areas of greatest
complementarity between these distinctive perspectives and methodological
strategies.
This goal may be contrasted with triangulation, a research strategy that
involves using several methods to reveal multiple aspects of a single empirical
reality (Denzin, 1978). A major assumption of the triangulation strategy is that

34

GALE MillER AND KATHR YN J. FOX OBSERVATION
sociological research is a discovery process designed to get at an objective truth
that may be systematized as a formal theory of social structure and proce.ss.
Triangulation assumes that looking at an object from more than one standpomt
provides researchers and theorists with more c�mpre�ensive knowle� ge
about the object. This approach also assumes that there IS an overwhelmmg
need for a single set of standards by which the methodological act c�n be
evaluated" (Denzin, 1978: 339). The bridging approach discussed here dIffers
from triangulation in its focus on using several methodological strategies to
link aspects of different sociological perspectives, not to dis.cover in�isputa�le
facts or to construct an all-encompassing theory about a smgle sOCIal realIty.
We set the stage by demonstrating the significance of discourse in ethno­
graphic studies, thereby uncovering the link between structural or insti.tutional
frames for interaction and the methods used to analyze the ways that dIscourse
shapes possibilities. Next, we treat the concerns, object!ves, .and techniq� es
of ethnomethodology, conversation analysis, FoucauldIan dIscourse studIes
separately in order to establish correspondence between them. In addition,
we will demonstrate the ways in which the questions that researchers ask
and the sites they choose for study establish some parallels as well between
these various methodologies. We conclude by discussing what can be gained
by bridging distinct approaches.
DISCOURSE A N D ETH N O GRAPHY
The significance of ethnomethodology, conversation analysiS, and Foucauldian
discourse studies for this chapter centers in their concerns for how language
and knowledge are related and are constitutive aspects of social life. Each
of these perspectives stresses how social life may be organized within multiple
social realities as well as how the realities are socially constructed through
our use of language, and the reflexivity of our accounts of social settings,
realities, and issues. The concept of reflexivity refers to the ways in which
our portrayals of social realities simultaneously describe and constitute the
realities (Garfinkel, 1967). Our descriptions of social realities, then, cannot
be separated from the objects, persons, or circumstances that they describe or
the languages that we use to describe them.
While informed by aspects of philosophy (particularly hermeneutics and
ordinary language philosophy), ethnomethodology and conversation analysis
orient toward and recast classic sociological issues (Hilbert, 1992). They also
emphasize how social realities are built from the "bottom uf (from ordin� ry
interactions to general social processes). Foucauldian dIscourse studIes,
on the other hand, are part of the philosophical movement sometimes called
poststructuralism (Eagleton, 1983) and other times postmodernism (Best
and Kellner, 1991). This approach also treats social realities as embedded
in generalized discourses which people enter into and u� e in � onducting th�ir
everyday activities and interactions. Indeed, FoucauldIan dIscourse studIes
might be characterized as moving from the "top down" (from culturally
36
B U I L D I N G BRIDGES
standardized discourses to the reality-constructing activities of everyday life).
Schutz (1970) was concerned with a similar problem, that is, the relationship
between the structural possibilities endowed by the "lifeworld" (or everyday
life) and the systems of relevance which characterize social groups. In other
words, meaning systems for groups may appear distinct from everyday life,
yet our relevance codes emanate from practical experiences (including
relationships to institutions) similar to the Foucauldian notion that discourse
conditions our possible understandings.
These perspectives are also similar in their treatment of empirical research
an analysis as interrelated. Ethnomethodology, conversation analysis, and
d
Foucauldian discourse studies are not offered as integrated, all-encompassing,
or grand theories of society (defined as an abstract structure or entity), but
as distinctive standpoints from which concrete, empirical aspects of social
life may be seen and analyzed. Their empirical focus is also reflected in the
differing methodological strategies associated with each perspective. The
strategies are designed to produce data that might be used to apply, extend,
and elaborate on issues that are central to the perspectives. While some
proponents of these perspectives and strategies describe their research as
inductive (e.g., Merry, 1990) because it involves careful analysis of data, it
is important to recognize that qualitative data - like other depictions of social
reality are social constructs. Thus, they are influenced by researchers'
assumptions about social reality and methodological practices.
Taken together, these perspectives provide qualitative sociologists with
interpretive resources for writing ethnographies of institutional discourse
(Miller, 1994). These ethnographies focus on the ways in which everyday life
is organized within, and through, language. They involve attending to both
the discursive categories and practices associated with social settings, and
how setting members use them (sometimes in distinctive ways) to achieve
their practical ends. Thus, social settings might be said to "provide" their
me�bers with discursive resources and opportunities for constructing a
vanety of social realities. Situationally "provided" discourses shape and guide
(but do not determine) what might be said in social settings (Silverman,
1987). Social realities are always locally constructed and contingent. They are
"built up" through setting members' organization and use of the discursive
resources and opportunities that are made available to them in concrete social
sett.ings. This analytic focus has, at least, two major implications for qualitative
SOCIOlogists' orientations to their research. First, ethnographic studies of insti­
tutional discourse need to be differentiated from qualitative studies that
focus on the distinctive values and perspectives of cultural and subcultural
groups. These studies provide readers with "insider" knowledge about how
c�lt�r.al and subcultural groups orient to social reality, and explain the social
slgmficance of their distinctive practices. Schwartz and Jacobs (1979) aptly
�l:aracterize this ethnographic approach as reality reconstruction, because
It IS concerned with accurately representing the meanings expressed by group
�embers. The ethnography of institutional discourse, on the other hand, better
fIts WI. t1un
' Schwartz and Jacobs' (1979) formal sociology category which
37

GALE MILLER AND KATHRYN J. FOX w OBSERVATION
focuses on setting members' interpretive and interactional competencies,
including those that are so taken for granted that members are unlikely to
mention them to one another or to qualitative researchers .
Second, ethnographies of institutional discourse extend the long-standing
emphasis on observational methods in qualitative sociology. The methods
include participant observation, various types of non-participant observation,
and the use of less obtrusive observational techniques (such as observa­
tion frombehind one-way mirrors or other "hidden" sites). Frequently (perhaps
usually), qualitative researchers combine these observational strategies with
other qualitative methods, such as interviews and life histories. Ethnographers
of institutional discourse also rely on observational methods (usually non­
participant observation) in conducting their research. But the focus of
discursively oriented ethnographers' observations is different from those
of other qualitative researchers.
One way of understanding this difference is by considering what it means
to study social settings versus social worlds. The latter research topic assumes
that everyday life is organized within relatively stable and integrated ways
of life (Unruh, 1983). Qualitative researchers of social worlds use observational
and related methods to identify and reconstruct the perspectives and patterns
of action and interaction that organize diverse social worlds. Discursively
focused research on social settings, on the other hand, emphasizes how social
realities are always under construction. It considers how setting members
continually assemble and use the interactional and interpretive resources
"provided" by social settings to construct defend, repair, and change social
realities. Hence the emphasis by discursively oriented ethnographers on
observing (directly, by means of audio and video recordings, and through
the careful reading of texts) the actual ways in which setting members
construct social realities by making sense of practical issues.
We elaborate on these issues in the next three sections by discussing
some of the major emphases of ethnomethodology, conversation analysis, and
Foucauldian discourse studies. The discussion is selective, emphasizing those
aspects of the perspectives that might be used to construct dialogue between
them. Later, we use this discussion to identify areas of complementarity
between the perspectives, and then consider how they might be bridged
through comparative research.
ETH N O M ET H O DOLOGICAL CONCERNS A ND STRATEG I ES
The ethnomethodological project focuses on the common-sense methods that
we use to make sense of our experiences and constitute social realities
(Garfinkel, 1967). The methods of special interest to ethnomethodologists are
the various interpretive procedures that we routinely use to classify aspects
of our experience and to establish connections between them. Smith (1978),
for example, takes an ethnomethodological stance in analyzing how persons
are assigned to mental illness categories by way of contrast structures. Contrast
38
BUILDING BRIDGES
structures are oppositional distinctions that cast some circumstances,
behaviors, or persons as normal, natural, or preferred and cast others as
abnormal, unnatural, or undesired. An example is the following statement
made by Angela about K.
(i) We would go to the beach or pool on a hot day,
(ii) I would sort of dip in and just lie in the sun
(iii) while K insisted that she had to swim 30 laps. (Smith, 1978: 43)
Angela's claims are expressed as straightforward, declarative stat�ments
that might be treated as descriptive of her own and K's behavior. For ethno­
methodologists, however, such descriptive practices are reality-creating
activities through which behaviors, circumstances, and persons are cast
as instances of cultural categories and may be assigned moral and political
significance. Also, two of the several assumptions about mental illness under­
lying Angela's description are that mental illness is a departure from what
might be called a normal state of mind, and that signs of mental health and
illness may be discerned from persons' behavior.
Foucault (1975) shares an analogous fascination with such categorization
of persons in many of his texts. His attraction to the memoir of Pierre Riviere
reflects a preoccupation similar to those of ethnomethodologists. The case
of Riviere, a man who murdered his family in the nineteenth century, includes
accounts by judges, villagers, the mayor, Riviere himself, among others.
Foucault (1975: x) sees the memoir as representing "a battle among discourses
an� thr�ug� di�courses." In other words, he examines the "tactics" used by
vanous InstItutIOnal actors to establish the memoir's "status as the discourse
of either a madman or a criminal" (xii).
Another ethnomethodologically interesting feature of Smith's account of
Angela involves its reflexivity and localness. Angela uses the account to
construct a world in which she and K are assigned distinct, contrastive, and
hierarchical positions and identities. Angela is positively positioned as normal
�nd K as mentally ill. The account is local because its meaning is inextricably
lInked to the practical circumstances in which it was voiced and interpreted
by others. One reason why local circumstances are important to ethnomethod­
ologist� is because they stress that social constructions of reality are always
potentially open to contest and change. Thus, Angela's description might
not always be treated as evidence of K's mental illness. Consider, for example,
how the meaning of Angela's description changes if we assume that K is
an Olympic swimmer in training or that swimming is part of K's rehabilitation
fro.m an accident. In these cases, the contrast between Angela's and K's orien­
tahon� to swimming might be taken as evidence of K's great (and admirable)
commItment to athletic excellence or to recovering from her accident.
Ethnom�thodologically informed studies of institutional settings also focus
?n the SOCIal and political contexts within which members use available
�nt�rpreti�e methods to construct social realities. These studies analyze how
SOCIal settmgs are organized as interpretive hierarchies (Dingwall et al., 1983),
39
GALE MILLER AND KATHRYN J. FOX " OBSERVATION
local cultures (Gubrium, 1989), or rhetorical domains (Miller and Holstein,
1995, 1996) within which some orientations to pra.ctical �ssues are us�ally
privileged over others. Thus some definitions of sOCl�l reahty are �ore hkely
to emerge in institutional settings than othe�s (�:lbnum, 19?�; M,I�ler" 199.1 � .
Similarly, as Foucault (1975) described the Slgmflcance of RIvIere s memOIr.
Documents like those in the Riviere case should provide material for a thor?� gh
-'xamination of the way in which a particular kind of knowledge (e.g. medlcme,
;sychiatry, psychology) is formed and acts in relation to institutions and the roles
prescribed in them. (xi)
Emerson and Messinger (1977), for exautple, analyze how trou?les. are
defined and responded to in human service and sOci.al c?ntrol or.g�m:atlOns
as a micropolitical process. They explain that orgamzatlOnal. offICIals usual
orientation to potential trouble definitions and reme d'les IS h'Ierarch'lca11y
arrayed along continua, ranging �rom the.n:o.st preferred .to the lea�t preferr�d.
In these settings, then, all pOSSIble defmltlO.ns ?f reaht� \and mte!pretlVe
methods) are not equally available to orgamzatlOnal offICials. As Em�r.son
(1969) shows in his study of decision making by juveni1� court off1c�als,
dispreferred definitions and remedies may, be so devalue� m these sett1�gs
that they are rejected in favor of more typIcal responses If only one settmg
member speaks against them.
While they analyze their data from a dlffe�ent sta�dpo�nt, ethnomethod-
. .
ologists usually observe social settings and mteractlOns m much the same
ways as traditional ethnographers. They observe and t�ke notes �bout the
everyday activities and relationships of th�ir �esearch s�bJects. Th� dlffere.nce,
as Coulon (1995) notes, does not lie in thelr fIeld techmques, but m the kmds
of questions that conventional and ethnomethodological ethnog�aphers. ask
about social settings and processes, and the types of data that th.elf questlOns
generate. As the above discussion suggests, ethnomethodologlsts are m?re
likely than other ethnographers to focus on the interpretive practi�es of �et�mg
members, a focus that requires that they attend to (a.nd. record m t�eIr fleld
notes) the details of setting members' interactions. A SImIlar concern IS central
to the methodological strategies of conversation analysis.
CONVER SATI O N ANALYT IC CONCER NS AND
Developed simultaneously with ethnomethodology: con�ersation an�lysis
focuses on the ways in which social realities and relatlOnshlps �re const1tu�ed
through persons' talk-in-interaction (Sa.ck� et a1., . 1974). !hlS pe�spect1ve
focuses on the social organization of talk-m-mteractlOn, the mteractlOnal and
interpretive competencies of the interactants, and how they collaborate to
construct social realities. While conversation analysts share ethnomethod­
ologists' interest in interpretive methods, they treat these methods as emergent
from the distinctive structure and processes of talk-in-interaction. In its most
basic form, talk-in-interaction is organized as sequentially organized turns at
40
B U I L D I N G BRIDGES
talk through which speakers reflexively construct a context for their inter­
actions as they go about the practical activities that make up the interaction.
These activities include displaying "appropriate" orientations to others' talk
as well as taking and finishing one's speaking turn in "appropriate"
ways.
Consider, for example, the following interaction occurring in a plea­
bargaining meeting, and involving a public defender (PD2) and district
attorney (DA3). Prosecuting and defense attorneys meet to negotiate the
charges to be made against defendants (in this case Delaney) and punishments
that will be given to defendants after they plead guilty to the charges.
1. PD2: Okay uh is there an offer in Delaney?
2. DA3: Yeah plea to Mal M ish and uh uhm modest fine and uh restitution
3. P D2 : Okay
4. (0.8)
5. PD2: Fifty dollars?
6. DA3: Yes. (Maynard, 1 984: 80)
This exchange displays several of the collaborative skills that we routinely
use in successfully interacting with others. First, notice how PD2 (line 1) opens
the interaction by stating "OK," thus marking off the previous discussion from
that which follows. PD2 then identifies the topic of the subsequent interaction
and invites a plea from DA3 by asking "is there an offer in Delaney?" DA3
accepts the request (line 2) by offering the charge of "Mal Mish" (malicious
mischief) and a punishment ("uhm modest fine and uh restitution") . PD2
continues to collaborate in the interaction by agreeing to the proposed charge
and punishment (line 3). The pause (line 4) might also be understood as
a collaborative act. It signals PD2' s and DA3' s readiness to move to a new topic
which PD2 raises by suggesting $50 as an appropriate amount for the fine
(line 5). DA3 then closes the interaction on line 6 by agreeing to the suggested
fine.
Conversation analysis is a context-sensitive approach to the study of reality
construction (Rawls, 1987). It focuses on the details and contingencies of social
interactions, and emphasizes how every social interaction is a distinctive
occasion for constructing social reality. But conversation analysts also analyze
social interactions as having elements that are context free because they are
evident in other social interactions. We see both context-sensitive and context­
free aspects of talk-in-interaction in the above exchange between PD2 and
DA3. While each move in the interaction is a local and collaborative accom­
plishment, PD2 and DA3 also orient to more general conversational practices.
Most obviously, they orient to the interaction as a set of turn-taking sequences
by waiting for, and then taking, their speaking turns in the interaction. They
also display general understandings about how conversational topics are
proposed, negotiated, and terminated through such mundane moves as stating
"Okay" and "Yeah" at the beginning of their speaking turns, and by using
the pause to manage a conversational shift.
41

GALE MillER AND KATHRY N J. FOX OBSERVATION
Conversation analysts share ethnomethodologists' interest in the distinctive
circumstances associated with talk and reality construction in institutional
settings. They emphasize how these interactions involve both context-fr� e
aspects of ordinary conversations, and how interactants assemble them m
distinctive ways to produce social contexts within which some interactional
patterns and social relationships are encouraged over others (Atkinson and
Drew, 1979; Drew and Heritage, 1992; Jefferson and Lee, 1992; Zimmerman
and Boden, 1991) . Further, the patterns and relationships associated with
institutional settings are unlikely to provide all members with equal oppor­
tunities and resources' for pursuing their interests in the interactions, thus
producing social conditions in which some definitions of social reality are
more likely to prevail than others.
For example, this focus is central to Perakyla and Silverman's (1991) analysis
of counseling sessions as communication formats within which setting
members take different social roles and positions. The communication formats
of most interest to Perakyla and Silverman (1991) are interviews (in which
counselors ask questions and patients answer them) and the information
delivery format (in which patients listen to the information and evaluations
conveyed by counselors). While different, both formats offer counselors
greater opportunities to express and pursue their interests, including their
preferred definitions of social reality. Indeed, we might extend this analysis
by considering how these communication formats are both contexts of, and
sources for, the distinctive professional authority and power exercised by
counselors in their interactions with clients.
Conversation analysts usually study social interactions by constructing
and analyzing transcripts made from audio and video recordings of social
interactions. The transcripts are fine-grained representations of the interactions
that often include notations indicating the length and placement of pauses,
simultaneous talk by interactants, speakers' intonation, words that are stressed
or elongated by speakers, and the direction of interactants' gazes. Because
interactants might take any of these aspects of social interactions into account
in responding to others' utterances or in moving the interactions in new
directions, they are relevant to conversation analysts analyses of how social
realities are interactionally constructed, sustained, and changed.
FOUCAULDIAN CONCERNS AND STRATEGI ES
Foucault uses the term discourse to analyze more than language. It also includes
the assumptions, logics, and modes of articulation associated with particular
uses of language. Discourses provide persons with coherent interpretive
frameworks and discursive practices for constructing different social realities
within which particular kinds of people reside, relationships prevail, and
opportunities are likely to emerge. We enter into discour��s as we go ���ut
the practical activities of our lives. The discourses are conditlons of possIbIl�ty
that provide us with the resources for constructing a limited array of SOCIal
42
E
BUILDING BRIDG S
realities, and make oth�r possibilities less available to us. Arney and Bergen
(1984) elaborate on thIS pOint in their discussion of medical discourse as
knowledge, power, and truth. They state:
It is m?re than just a set of facts known by physicians and embodied in a
professlOnal, specialized, inaccessible language. The medical discourse is a set of
rules that enables facts to become facts for both physicians and patients. It is
a s�t of rules that Covers not only what is important to doctors but also what
patIents can speak about as important. Knowledge is power precisely because the
knowl�dge embedded in the medical discourse supplies rules by which patients
ascertal� when they �re s�eaking true about the self and when they are speaking
about t�ngs that are nnagmary, Kn�wledge tells the person what is important and
,
not fanCIful about hIS or her expenence of illness and p atienthood, (Arney and
Bergen, 1984: 5) ,
In a similar vein, Ainsworth-Vaughn (1998) demonstrates the more micro
aspe �ts of ,negotiation in medical encounters. As opposed to studies that
proVIde eVIdence of the power over patients that doctors possess and wield,
�he anal�zes th,e ways that doctors and patients claim power, or struggle over
�t rheton,cally, m exc�anges . . Cer,tainly, �uch an inquiry suggests that power
IS a dommant theme m medIcal mteractIOns; however, Ainsworth-Vaughn's
study, exa,mines the small ways that utterances reveal grabs for power in
the mmutla of talk. Thus, power is produced within the struggle over it.
It matters, then, which discourse we enter into to organize and make sense
of th� practical issues emergent in our lives. For example, the discourse of
law IS only one of several discourses available in contemporary Western
cultures for makI, �g sense of, and responding to, interpersonal and intergroup
conflI, �ts, C?thers mclude moral, mediative, and therapeutic discourses, each
of whIch mvolves assumptions, categories, logics, claims, and modes of
articulation ,that differ from those making up legal discourse. While politically
consequentlal, our entrance into discourses is usually experienced as un­
remarkable because we associate different discourses with different kinds
of setting� . �hus, discourses might be said to have their own social settings,
although It IS uncommon for only one discourse to be available in a social
setting,
Conley and O'Barr (1990) show, for example, that while small-claims courts
a�e domin�te ? by le�al discourse, judges and litigants occasionally organize
dispute � wIt�m t�e dlscOl�rse of relationships, The latter discourse emphasizes
t�e SOCIal hIS tones .of dIsputants, and the distinctive (often extenuating)
Clrcu,mstances assocIated with their disputes, The discourse of relationships
also mvolves a distinctive arrangement of power and authority within the
co�rtroom. That is, the disputants (not judges and lawyers) possess authori­
���ISlvedIScourse.
.knowledge about the so�ial histories,and circumstances emphasized in
Others are restncted to helpmg the disputants tell their stories,
and develop mutually agreeable solutions to their disagreements.
A�alyzing the availability of multiple discourses in social settings also raises
questIons about the discontinuities within, and between, the discourses, This
43

GALE MILLER AND KATHRYN J. FOX " OBSERVATION
concern is perhaps most evident in Foucault's (1970 , 1972) approach to
historical change, which he analyzes as filled with radic al disjunctures or
replace old ones. While
ruptures that occur when new discourses emerge and l settings involving
not directly applicable to qualitative studies of socia urse studies is still
more limited time spans, this theme in Foucauldian disco us of the possibilities
relevant to qualitative research. At the least, it remindsdisco ntinuities might
for discursive discontinuities in social settings. Thesemem move between
bers
be observed, for example, in the ways in which setting issue s of everyday life,
different discourses in dealing with the practical disco produce
and when setting members operate within availableissues. urses to
unanticipated and atypical orientations to practical are analyses of historical
While most applications of Foucault's perspective
studies show that these
texts, Merry's (1990) and Conley and O'Barr's (1990)vatio techniques, and
issues may also be studied by using interviews, obser Foucnal auldi an disco�rse
tape-recorded data. Whatever the form of the data,culturally stand ardized
studies involve treating the data as expressions of l settings. Foucauldian
discourses that are associated with particular sociaassumptions, categories,
researchers scrutinize their data, looking for related urses. They also analyze
logics, and claims - the constitutive elements of disco in social settings, how
how different (even competing) discourses are present , the political positions
related social settings may involve different discourses discursive practices
of setting members within different discourses, and the ete issues,
used by setting members to articulate and apply discourses to concr
persons, and events.
ESTABLISH I N G
Despite its limitations, the above discussion of ethnomethodology, con­
versation analysis, and Foucauldian discourse studies provides us with
a beginning for identifying and elaborating on areas of complementarity
between the perspectives. These areas are, of course, easier to see in comparing
ethnomethodology and conversation analysis because they are informed
by the same intellectual traditions, address similar questions, and focus on
similar aspects of everyday life. They are similar, for example, in their concern
for how social realities are "built up" and sustained. Unlike Foucauldian
scholars who focus on the general categories, practices, and logics of histori­
cally emergent discourses, conversation analysts emphasize the interpretive
and interactional methods (both context sensitive and context free) that people
in concrete social situations use to construct realities. The next two sections
discuss two related strategies for linking ethnomethodology and conversation
analysis, on the one hand, and Foucauldian discourse studies, on the other .
44
G
B U I L D I N G BRID ES
M ETHODS AS ANALYTIC O .-. " ,-,, \.,
Ethnomethodologists, conversation analysts, and Foucauldian scholars orient
to, and rely upon, empirical data to develop their perspectives. The data
analyz.ed by thes� theorists are not merely materials for illustrating aspects
of �helr perspectives. The analyses are theory-constructing activities in
WhiCh data are a central focus. To be sure, the data that these theorists usu­
ally. analyze are different, but the differences are at least partly matters of
�hOlce, not absolute necessities. Indeed, the literature of qualitative sociology
mcludes several ex�mples of ho,: ethnomethodology, conversation analysis,
and FoucauldIan . discourse studies can be done by using "unconventional"
methods and data.
H�lst:in' � (1993) study of legal proceedings concerned wjth the involuntary
hOspitalIzatIon of persons diagnosed as mentally ill, for example, shows how
conversation analytic concerns can be successfully addressed by Ilsing data
collected through observational methods . This analysis was successful because
Holstein (1993) brought an appreciation of the significance of talk-in-interaction
to his fieldwork. While these data are not so richly detailed as those that
might be gleaned from audio and video recordings, they are sufficient for his
�nalytic tas�s and incl:rde co�textual information that is not always present
111 conversatIOn analytic studIes based only on mechanical recordings.
Also, McHoul's (1982) and Silverman's (1975) research illustrates ho w
ethnomethodologists analyze written texts. While he does not cite Foucault,
McHoul (1982: x) casts his project in a Foucauldian language when he states,
The discursive order . . . produces every possible version of 'the social' . . . and that
will be �o whether it is ethnomethodological discourse that is in question or one
of the dIscourses that ethnomethodology would preferably take as its 'object'.
Conley and O'Barr (1990), Merry (1990), Miller (1991), and Silverman
(1987) address Foucauldian issues by using ethnographic and conversational
data . While different in their empirical and analytic aims, each of these
studies considers how social life is organized within institutional discourses,
and how knowledge and power are implicated in them. They advance the
Foucauldian project by linking it to qualitative researchers' interests in
the social organization of everyday life.
Taken together, these developments in qualitative research suggest that
data are not always a problem in creating dialogue between ethnomethod­
ology, conversation analysis, and Foucauldian discourse studies. The analytic
concerns . of the p :rspectives can be successfully addressed by analyzing
several dIfferent kll1ds of data. Of course, it is easy to take this claim too far,
because these perspectives are not compatible with any kind of data. At
t�e very le�st, conversation analysis requires data that are amenable to sequen­
hal analY �Is, ethn�methodology requires data that might be analyzed as
mund�ne ll1terpretlve methods, and Foucauldian analysis requires discursive
data. 1 hese data can be generated within the same or linked qualitative studies,
45
GALE MILLER AND KATHRYN J. FOX OBSERVATION
however, by combining two or more data generating techniques (such as
observations, tape recordings of social interactions, and textual analysis).
Another promising approach to building bridges linking ethnomethodology,
conversation analysis and Foucauldian discourse studies is comparative
qualitative research.
There are several recent examples of how qualitative researchers have
comparatively analyzed institutional settings, and addressed analytical
issues in the process. Gubrium's (1992) ethnographic studies of two family
therapy clinics is an example . While presumably dealing with similar prob­
lems, Gubrium's (1992) analysis shows how the therapists defined and
responded to their clients' troubles in very different ways, in one case treating
them as family system problems and in the other as emotional troubles.
Further, because Gubrium's (1992) fieldwork was informed by the ethno­
methodological perspective, his study provides detailed information about
the mundane interpretive and interactional practices through which the
therapists entered into, and operated within, these discourses to produce
organizationally preferred trouble definitions and remedies.
Miller and Silverman's (1995) comparative study of an AIDS counseling
center in London and family therapy clinic in the United States is an example
of how a comparative strategy may be implemented to address ethno­
methodological, conversation analytic, and Foucauldian concerns . In this
case, the researchers collected their data independent of one another, and
then collaborated by analyzing the continuities and discontinuities in the data .
The study illustrates how conversation analytic and Foucauldian approaches
to troubles talk may be bridged through the use of qualitative data . They
pursue the latter goal by treating their data as venues for exploring situational
and transsituational aspects of counseling discourse, settings, and practices.
Miller and Holstein (1995, 1996) take a different approach to comparative
research and analysis in their study of conflict emergence and dispute
processing in one human service organization. The study is based on extensive
observations of everyday life in the organization and analysis of audio tapes
of legal proceedings concerned with dispute resolution. The analysis �ocuses
on the ways in which conflicts and disputes are differently orgamzed as
they are considered within different dispute domains which Miller and
Holstein (1995, 1996) analyze as made up of the typical assumptions, concerns,
vocabularies, and interactional practices associated with different social
settings. This study, then, offers a distinctive view of the conditions of
possibility associated with the evolution of disputes in one organization.
Miller and Holstein (1995, 1996) consider how organizational settings are
linked to form an ecology of knowledge and power, and how outside parties
(legal officials associated with a different government agency) sometimes
become involved in ongoing disputes. The study considers the ways in which
disagreements are changed as they are configured and reconfigure? wit�in
the conditions of possibility provided by each dispute domain. WhIle mam­
taining an interest in Foucauldian issues, Miller and Holstein (1995, 1996)
also stress ethnomethodological themes. They analyze, for example, how
46
B U I L D I N G BRIDGES
members of these settings use available interpretive and interactional
resources to artfully pursue their interests and, based on members' actions,
how disputes sometimes move in unanticipated directions.
A final comparative research and analytic strategy involves longer term
(even longitudinal) studies of one or a few settings. The focus here is on how
discourses and their associated interpretive and interactional practices change
over time . This strategy might be useful in observing the disjunctures or
ruptures that Foucault emphasizes in his historical studies of social change,
but they may also provide qualitative researchers with information about
the ways in which discourses, settings, and related interpretive and inter­
actional practices evolve over time. While less dramatic than studies of
radical disjunctures, studies of the evolution of discourses and their related
settings and practices provide insights into the potentially unstable and
changing character of language, culture, and institutions ( see, for example,
Miller, 1997) .
In sum, these studies remind us of the complexities of everyday life and
how agency and constraint are simultaneously implicated in it. An exclusive
fo�u� on either side of this dichotomy is inadequate, since everyday life is lived
withm cultu �ally s.tandardized discourses and the discourses are changed
by the ways m WhICh we use them. While ethnomethodo]ogy, conversation
analysis, and Foucauldian discourse studies are distinctive approaches to these
issues, comparative qualitative research that bridges them provides analysts
with conditions of pOSSibility for artfully extending and displaying (in concrete
detail) the importance of their insights.
QUESTIONS A N D SITES AS ANALYTIC B R I DG ES
Qualitative research and analyses are as much about asking questions as
providing answers. Thus, asking questions that address themes that are
part of, or imp�ied by, two or more perspectives is a form of analytic bridg­
mg. The questIOns allow qualitative researchers to focus on aspects of the
perspectives that are - at least potentially - compatible . For example, we might
ask, "How do ethnomethodological, conversation analytic and Foucauldian
re�earch strategies address aspects of the agency-constraint debate in the social
SCIences?" The debate focuses on the extent to which social realities and actions
are products of individual initiative or are shaped by larger social forces .
Ethnomethodologists and conversation analysts might be seen as stressing
age.ncy ov�r constraint because they focus on the local and artful ways
:-vhIch settmg members assemble and use available interpretive resources
111 formulating their understandings of, and responses to, practical issues.
Foucauldian discourse studies, on the other hand, might be interpreted
as stressing constraint over human agency because they focus on the ways in
:-vhich
. the assumptions and interpretive procedures used by setting members
I� constructing social realities are provided by the culturally standardized
dIScourses that predominate in social settings.
47
GALE MILLER AN D KATHRYN J. fOX OBSERVATION
This interpretation becomes problematic, however, when we consider
ethnomethodologically informed and conversation analytic studies concerned
with institutional settings and talk. These studies provide both data and
interpretive frameworks for analyzing the practical constraints which insti­
tutional actors take into account in pursuing their interests in social settings,
and for assessing the micropolitical advantages enjoyed by some setting
members in pursuing their interests in institutional settings. Similarly, agency
is an aspect of Foucauldian inspired qualitative studies such as Conley and
O'Barr's (1990) which consider how institutional actors ("artfully") enter
into available discourses and shift from one discourse to another in dealing
with practical issues. Agency and constraint are not mutually exclusive issues
in these studies, then, but coterminous aspects of the settings under study and
appropriate topics for study in their own right.
A promising area for reconsidering the agency--constraint debate involves
recent qualitative studies of gender, race, and class. The issues are central
to Foucauldian studies of discourse, knowledge, and power. Foucauldian
scholars treat gender, race, and class as aspects of unequal power relations.
Foucault analyzes inequality as a product of people's everyday social practices
and activities, that is, as power relations. Within our power relations, we
discursively construct realities that justify and sustain gendered, racialized,
and class-based inequalities. This approach to discourse and power resonates
with Smith's (1987, 1990) institutional ethnography approach to these issues.
Smith (1990: 80) states that structures of power and privilege do not exist
separate from people's actions. Rather, they are produced and sustained
through the "mobilization of people's concerted actions."
Smith's approach to gender, race, and class opens the possibility for
building bridges between Foucauldian discourse studies, ethnomethodology,
and/ or conversation analysis through qualitative research. The studies might
focus on the mundane interpretive and conversational practices used by
individuals and groups to construct, sustain, and sometimes resist differ­
ences in power and privilege. Smith's previously discussed analysis of
contrast structures and mental illness is an example of such bridging.
For Smith, these contrast structures and related interpretive practices are
aspects of larger political relations that express and sustain psychiatric
power and patriarchy.
Another example of how such bridges can be built is Miller's (1993)
postructuralist and feminist analysis of part of an interview with Alice Dunn
and her husband, Michael Dunn (first published in Brannen and Moss, 1987) .
Central to Miller's analysis is the Foucauldian assumption that power infuses
all talk and therefore all talk is political. The segment under analysis began
when the interviewer asked Mrs. and Mr. Dunn about the importance of
Mrs. Dunn's earnings for the household budget (Brannen and Moss, 1987: 159).
Their interaction turned on whether Mrs. Dunn's earnings were necessary
for the household or were used to purchase "luxuries." Mrs. Dunn first
stated, "I think it's quite necessary," but in the end, she says, "Yes, I suppose
it is for luxuries." (Brannen and Moss, 1987: 89).
48
B U I L D I N G BRIDGES
Miller draws upon and critiques aspects of ethnomethodology and con­
versation analysis in analyzing how this interaction depoliticizes Mrs. Dunn's
initial portrayal of her earnings as necessary to maintaining the household.
Foucault might say that the interaction shows how cultural discourses
encourage some portrayals of social reality and discourage other portrayals.
Miller (1993: 167) also suggests how ethnomethodologists and conversation
analysts might "broaden" their analyses "to include large-scale discourses
that influence the ways interaction is accomplished" while not reducing
"speakers to the puppets of these macrolevel forces." For Miller, a useful
research strategy for extending ethnomethodology and conversation analysis
involves noticing and analyzing the "underdog" methods and strategies
that marginalized groups and individuals use in countering claims based on
dominant discourses, as well as how "top dogs" sometimes use these methods
and strategies for their purposes. '
A related example of how qualitative researchers might study agency and
constraint as coterminous aspects of everyday life is Chase's (1995) analysis
of the work narratives of women school superintendents in the United
States. Chase displays an ethnomethodological awareness in describing the
social contexts of her interviews with the superintendents. She stresses that
it is not enough to pay attention to what the superintendents say about
their lives and experiences. Chase (1995: x) adds that qualitative researchers
must also pay attention to how interviewees tell their life stories "by attending
to the cultural, linguistic and interactional contexts and processes of story­
telling."
Chase focuses much of her analysis on two major cultural discourses
or counternarratives used by the superintendents in telling their life stories.
One discourse stressed the many accomplishments achieved by the super­
intendents in their professional lives. The other discourse consisted of stories
about gender, racial, and ethnic discrimination encountered by the super­
intendents and by other women in pursuing their careers. Chase notes that
the women in her study were confident and adept at using both discourses
in talking about their lives. This changed, however, when the superintendents
were asked to link the discourses. Chase (1995: 11) analyzes this request
as creating a discursive disjunction because "talk about professional work
and talk about inequality belong to two different discursive realms, two
conflicting vocabularies for articulatin g experience, two different ways
of talking about oneself."
Chase further develops this theme by analyzing the various narrative
strategies that the superintendents used to manage the disjuncture between
the discourses of professional success and social inequality. Her analysis
constructs a bridge between Foucauldian discourse studies and ethnomethod­
?logy (and perhaps conversation analysis) by displaying the interactional and
I�t�rpretive skills that the superintendents used in managing this discursive
dISJ uncture. These skills are often taken for granted in Foucauldian and
?ther macroscopic studies of power, knowledge, and language. Further, the
ll1terviewees' efforts to manage the disjuncture display discursive problems
49
GALE MILLER AND KATHRYN J. FOX OBSERVATION
and competencies that sometimes are overlooked in ethnomethodological and
conversation analytic studies.
Another example of how analytic bridges may be built by asking questions
is Jackson'S (2001) recent ethnography of Harlemworld. Jackson takes a
synthesizing approach to analyzing Harlemworld as a place within New
York City and as a symbol of African American h�stor� and cult.ure. He dr�ws
upon Foucauldian discourse studies in developmg hIS ,analyttc perspectJ ve.
But his analysis also includes themes that resonate WIth aspects of ethno­
methodology. The themes involve Jackson's analysis of race and clas � as
socially constructed and performed realities. Put differently, Jackson defmes
race and class as activities that members of Harlemworld do . Jackson
(2001: 4) explains that
many African Americans have decidedly performative notions of soci�l identi�y.
Class position is glimpsed through interpretations of e�eryday behavIO.rs. RacIal
identity is predicated on perceptions of particular social actIOns and IS shored
up with recourse to specific kinds of activities. Racial "location" is not contingent
solely on one-drop rules or degrees of skin pigmentati �n. Sociall� meaningful
.
identifications are partially derived from observable behavIOrs, practlces and SOCIal
performances.
Jackson's study has important implications for the ethnomethodologi.cal
study of race and class. The study provides a starting point for studymg
the mundane methods (performances) that people use to cast themselves
as members of particular social classes and races, and the methods used by
others to "read" the performances as evidence of social class and racial mem­
bership. Jackson'S ethnography also details the varia.bility . of pe? ple's
social class and racial performances (and claims) across SOCIal settmgs . Fma.lly,
Jackson's study might be read as an instruction to ethnomethodoiogists
about how they might analyze mundane reality construction as embodied
activity. . constructlOn.
Barrett (1996), too, demonstrates the embodiment of reahty
in his study of schizophrenic "cases." Barrett (1996: 19) analyzes how ':ps�­
chiatric teams" (made up of psychiatrists, social workers, and psychIatnc
nursing staff) variously construct a single schizophrenic patient differently
as a "segmented case," a "fully worked-up case," or a "whole person". Barrett
(1996: 19) fuses phenomenology and Foucauldian discourse studies insofar
as he analyzes the ways th� t "psy�hi�tr.ic disco�rse" character.ize� �ati;nts
as "variously endowed With SUbjectIVIty or dIvested of subJectlvlty . . In
contrast to Jackson'S study across sites, Barrett looks at the vantage pomts
and positions of actors within a single institution, a�d �ow the ac�ors' engage­
ment with the institution and patients frames theIr mterpretatlOn of a case
or illness.
50
B U I L D I N G BRIDGES
CONCLUSION
One advantage of the bridging metaphor is that it avoids the imagery of
totalizing synthesis in which the distinctive themes and contributions of two
or more perspectives are de-emphasized (even lost) in the interest of devel­
oping grand theoretical schemes. The bridging approach offered here seeks
to make different perspectives mutually informative, not to obscure or deny
their distinctive features . To that end, we conclude by discussing two general
implications of our approach to paradigm bridging through qualitative
research.
The first implication involves the selection of minimally compatible
perspectives. All sociological perspectives are not equally amenable to the sort
of linkage that we describe here. Thus, an early task of qualitative researchers
involves specifying the conditions of compatibility between the perspectives
which they wish to link through their research, such as we have done in
noting the complementary emphases in ethnomethodology, conversation
analysis, and Foucauldian discourse studies . How they differ on these issues
is also relevant, but discussion of the differences is only possible after areas
of compatibility have been specified.
Second, qualitative researchers' bridging projects must include the analysis
of data about research sites . While the qualitative research tradition includes
numerous and significant theoretical contributions, these developments
should not be separated from the empirical focus of the tradition . This state­
ment may be a source of controversy for some readers who properly
reject past claims by some qualitative researchers that their data are objective
facts that "speak for themselves," and should be treated as authoritative
adjudicators of theoretical disagreements about the nature of the "real" world .
This position neglects the ways in which qualitative data are themselves social
constructions that reflect the assumptions and practices of the researchers
who produced them.
Acknowledging that qualitative data are social constructions, however,
does not render them theoretically useless or irrelevant (see, for example,
Miller and Fox, 1999). Rather, acknowledgment recasts them as aspects of
a distinctive discourse that treats the practices of everyday life as worthy topics
of analysis . It also reminds qualitative sociologists that while theory is - by
definition - abstract, it should also speak to issues that are recognizable as
features of persons' everyday lives and social worlds .
Recommended re::lld irlfl
51
 B U I L D I N G BRIDGES
GALE MILLER AND KATHRYN J. FOX OBSERVATION

Conley, John M. and William M. O'Barr (1990) Rules Versus Relationships: The
Ethnography of Legal Discourse. Chicago: University of Chicago Press.
Coulon, Alain (1995) Ethnomethodology. Thousand Oaks, CA: Sage.
Dcnzin, Norman K. (1978) Sociological Methods: A Sourcebook. New York: McGraw-·
Hill.
Dingwall, Robert, John Eekelaar and Topsy Murray (1983) The Protection of Children:
State Intervention and Family Life. Oxford: Basil Blackwell.
Drew, Paul and John Heritage. eds. (1992) Talk at Work: Interaction in Institutional
Settings. Cambridge: Cambridge University Press.
Dreyfus, Hubert L. and Paul Rabinow (1982) Michel Foucault: Beyond Structuralism
and Hermeneutics. Chicago: University of Chicago Press.
Eagleton, Tcrry (1983) Literary Theory: An Introduction. Minneapolis: University of
Minnesota Press.
Emerson, Robert M. (1969) Judging Delinquents. Chicago: Aldine.
Emerson, Robert M. and Sheldon L. Messinger (1977) "The Micro··Politics of Trouble,"
Social Problems 25: 121--35.
Foucault, Michel (1970) The Order of Things: An Archaeology of the Human Sciences. New
York: Vintage Books.
- (1972) The Archaeology of Knowledge and the Discourse on Language, trans. A.M.
Sheridan Smith. New York: Harper & Row.
--

--- (1980) Power/Knowledge: Selected Interviews and Other Writings 1 9 72-1977 ed. Colin
Ainsworth-Vaughn, Nancy (1998) Claiming Power in Doctor-Patient Talk. New York:
Oxford University Press.
Arney, William Ray and Bernard J. Bergen (1984) Medicine and the Management ofLiving:
Taming the Last Great Beast. Chicago: University of Chicago Press.
Atkinson, J. Maxwell and Paul Drew (1979) Order in Court: The Organization of Verbal
Interaction in Judicial Settings. Atlantic Highlands, NJ: Humanities Press.
Atkinson, J. Maxwell and John Heritage, eds. (1984) Structures of Social Action: Studies
in Conversation Analysis. Cambridge: Cambridge University Press.
Barrett, Rob (1996) The Psychiatric Team and the Social Definition of Schizophrenia:
An Anthropological Study of Person and Illness. Cambridge: Cambridge University
Press.
Best, Steven and Douglas Kellner (1991) Postmodern Theory: Critical Interrogations. New
York: The Guilford Press.
Boden, Deirdre and Don H. Zimmerman, eds. (1991) Talk and Social Structure:
Studies in Etlmomethodology and Conversation Analysis. Berkeley, CA: University of
California Press.
Brannen, Julia and Peter Moss (1987) "Dual Earner Households: Women's Fina�cial
Contribution After the Birth of the First Child," pp. 75-95 in Give and Take in Famzlzcs:
Studies in Resource Distribution, eds. J. Brannen and G. Wilson. London: Allen and
Unwin.
Button, Graham and John. R. E. Lee, eds. (1987) Talk and Social Organisation. Clevedon,
UK: Multilingual Matters.
ol
Chase, Susan E. (1995) Ambiguous Empowerm ent: The Work Narratives of Women Scho
of Massachu setts Press.
Superintendents. Amherst, MA: University
cd. (1975) I, Pierre Riviere, having slaughtered my mother, my sister, and my brother
. . . : A Case of Parricide in the 1 91h Century, trans. Frank Jellinek. Lincoln, NE:
University of Nebraska Press.
Gordon. New York: Pantheon Books.
Garfinkel, Harold (1967) Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice
Hall.
Gubrium, Jaber F. (1989) "Local Cultures and Service Policy," pp. 94-1 12 in The Politics
ofField Research: Sociology Beyond Enlightenment, cds. J .F. Gubrium and D. Silverman.
London: Sage.
-- (1992) Out of Control: Family Therapy and Domestic Disorder. Newbury Park, CA:
Sage.
Heritage, John (1984) Garfinkel and Ethnomethodology. Cambridge: Polity Press.
Hilbert, Richard A. (1992) The Classical Roots of Ethnomethodology: Durkheim, Weber,
and Garfinkel. Chapcl Hill, NC: University of North Carolina Press.
Holstein, James A. (1993) Court-Ordered Insanity: Interpretive Practice and Involuntary
Commitment. New York: Aldine de Gruyter.
Jackson, John L., Jr. (2001) Harlemworld: Doing Race and Class in Contemporary Black
America. Chicago: University of Chicago Press.
Jefferson, Gail and John Lee (1992) "The Rejection of Advice: Managing the Problematic
Convergence of a 'Troubles-Tclling' and a 'Service Encounter' ," pp. 521-48 in Talk
at Work: Interaction in Institutional Sef:tings, eds. Paul Drew and John Heritage.
Cambridge: Cambridge University Press.
Lindstrom, Lamont (1990) Knowledge and Power in a South Pacific Society. Washington,
DC: Smithsonian Institution Press.
Maynard, Douglas W. (1984) Inside Plea Bargaining: The Language of Negotiation. New
York: Plenum Press.
McHoul, A.W. (1982) Telling How Texts Talk: Essays on Reading and Ethnomethodology.
London: Routledge & Kegan Paul.
Mehan, Hugh and Houston Wood (1975) The Reality of Ethnomethodology. New York:
John Wiley & Sons.

53
52
GALE MILLER AND KATHRYN J. FOX OBSERVA TION
B U I L D I N G BRIDGES
i!

Merry, Sally Engle (1990) Getting Justice and Getting Even: Legal Consciousness Among Zimmerman, Don H. and Deirdre Boden (1991) "Structure-in-Action: An Introduction,"
Working-Class Americans. Chicago: University of Chicago Press. pp. 3-21 in Talk and Social Structure: Studies in Ethnomethodology and Conversation
Miller, Gale (1991) Enforcing the Work Ethic: Rhetoric and Everyday Life in a Work Incentive AnalysIs, eds. D. Boden and D.H. Zimmerman. Berkeley, CA: University of
Program. Albany, NY: SUNY Press. California Press.
-- (1994) "Toward Ethnographies of Institutional Discourse: Proposal and
Suggestions," Journal of Contemporary Ethnography 23(October): 280-306.
(1997) Becoming Miracle Workers: Transformations in the Discourse and Practice of Brief
Therapy. Hawthorne, NY: Aldine de Gruyter.
Miller, Gale and Kathryn J. Fox (1999) "Learning From Sociological Practice: The
Case of Applied Constructionism," The American Sociologist 30 (Spring): 55--74.
Miller, Gale and James A. Holstein (1995) "Dispute Domains: Organizational Contexts
and Dispute Processing," The Sociological Quarterly 36(1): 37-59.
-- (1996) Dispute Domains and Welfare Claims: Conflict and Law in Public Bureaucracies.
Greenwich, CT: JAI Press.
Miller, Gale and David Silverman (1995) "Troubles Talk and Counseling Discourse:
A Comparative Study," The Sociological Quarterly 36(4): 725-47.
Miller, Leslie J. (1993) "Claims-Making from the Underside: Marginalization and
Social Problems Analysis," pp. 153-80 in Constructionist Controversies: Issues in
Social Problems Theory, eds. G. Miller and J.A. Holstein. Hawthorne, NY: Aldine de
Gruyter.
Periikylii, Anssi and David Silverman (1991) "Reinterpreting Speech- Exchange
Systems: Communication Formats in AIDS Counselling," Sociology 25(4):
627-51 .
Rawls, Anne Warfield (1987) "The Interaction Order Sui Generis: GoHman's
Contribution to Social Theory," Sociological Theory 5(Fall): 136-49.
Sacks, Harvey (1992) Lectures on Conversation: Volumes I & II, ed. Gail Jefferson.
Oxford: Blackwell.
Sacks, Harvey, Emmanuel A. Schegloff, and Gail Jefferson (1974) "A Simplest
Systematics for the Organization of Turn-Taking for Conversation," Language 50:
696-735.
Schutz, Alfred (1970) Reflections on the Problem of Relevance, ed. and trans. Richard
M. Zaner. New Haven, CT: Yale University Press.
Schwartz, Howard and Jerry Jacobs (1979) Qualitative Sociology: A Method to the
Madness. New York: The Free Press.
Shumway, David R. (1989) Michel Foucault. Boston: Twayne.
Silverman, David (1975) Reading Castaneda: A Prologue to the Social Sciences. London:
Routledge & Kegan Paul.
-- (1987) Communication and Medical Practice: Social Relations in the Clinic. London:
Sage.
Smith, Dorothy E. (1978) '''K is Mentally Ill': The Anatomy of a Factual Account,"
Sociology 12(1): 23-53.
--�- (1987) The Everyday World as Problematic: A Feminist Sociology. Boston: Northeastern
University Press.
-- (1990) The Conceptual Practices ofPower: A Feminist Sociology of Knowledge. Boston:
Northeastern University Press.
Unruh, David R. (1983) Invisible Lives: Social Worlds of the Aged. Beverly Hills, CA:
Sage.
Zimmerman, Don H. (1969) "Record-Keeping and the Intake Process in a Public
Welfare Agency," pp. 319-54 in On Record, ed. S. Wheeler. New York: Russell Sage
Foundation.

54
55
 ANALYSI N G DOCUMENTARY REALITIES

whether our own near-at-hand, or one to which we are strange and distant
incorporates an appreciation of those processes and products of self­
description . Consider, for instance, an ideal-typical organization. It goes
virtually without saying that this quintessentially modern kind of social
formation is thoroughly dependent on paperwork. Administrators, accoun­
tants, lawyers, civil servants, managers at all levels, and other experts or
specialist functionaries are all routinely, often extensively, involved in the
production and consumption of written records and other kinds of document.
If we wish to understand how such organizations work and how people
work with/in them, then we cannot afford to ignore their various activities
as readers and writers. Moreover, if we wish to understand how organizations
function on a day-to-day basis, then we also need to take account of these
routine tasks and roles of recording, filing, archiving and rftrieving informa­
Paul Atkinson and Amanda Coffey
tion. Indeed, the collective organization of work is dependent on the collective
memory that such written and electronic records contain.
In addition to these familiar record-keeping tasks, organizations also
produce significant documents of other kinds, including a variety of materials
concerned with their self-presentation. These might involve annual reports,
I NTRODUCTION: DOCU M ENTARY
prospectuses, financial accounts and the like. Many, though by no means all
of those documents, are produced for externat even public, consumption.
A significant amount of contemporary ethnogra p�ic fi�ldwc�rk takes place They may be among the methods whereby organizations publicize themselves,
in literate societies, in organizational or other settm gs m WhICh documents compete with others in the same marketplace or justify themselves to clients,
ogra phic fieldwork was his­ shareholders, boards of governors or employees. In the contemporary world,
are written read stored and circulated. Ethn we should also include electronic and digital resources among the ways in
torically co�ceiv�d and developed for research by SOCIa in esse �tially oral setti.ngs
�nthropol�gl�ts, which documentary realities are produced and consumed. Organizations,
not only in non-literate societies, often studied advancedl lIterate sOCle tles. for example, produce websites, promotional videos and similar artefacts.
but also in oral cultures or subcultures in more cond ucted in settin gs that are These are all among the techniques and resources that are employed to create
In contrast, contemporary fieldwork is often l actors. While such do�u­ versions of reality and self-presentations. Over and above these institutional
themselves documented by the indigenous socia it is important to. recogm�e documents there are also documentary records that embody individual
mentary work is rarely 'ethnographic' in itself,are self-documentmg. In t.hIS actions, interactions and encounters. People-processing professions, for
the extent to which many cultures and settin ical consgs
nces of �tudYI.ng example, routinely compile documents and records of professional--client
chapter we consider some of the method?lo � someeque of the ways m whIch interactions, in the production of medical records, case notes in social work,
ate
'documentary' societies and cultures. We mdlcstudy of dacumentary re�lit�es
qualitative field researchers can set about �he. the fabn. c of everyday socIal
school records and so on . These written records can be used to inform future
action, and are themselves fed into the more formal recording (and docu­
and the location of documentary work wlthm�itative res archers conti ue to mentary) mechanisms of official statistics, performance indicators, efficiency
life. It is important to do so because many qua rate SOCla � .n
hte � worl ds as If they league tables and similar constructs.
produce ethnographic accounts of complex, . shed Studl S of, for exampl�, The purpose of these introductory paragraphs is not simply to list a
were entirely without writing or texts. Man y pubh �
orga nizat ional nd eve .n educ atIOn al or acad emIC few indicative types of documents or to begin to outline what some of their
ts and. ot�er
occupatio nal, profe ssion al, � functions might be. Rather, it is to remind us of the pervasive significance
Id of wntt en docu men
settings are implicitly represented as deVO therefore, always do JustIce of documentary records, written and otherwise, in contemporary social (and
forms of textual recording. Such accounts do not, is necessary to redress t�e hence research) settings. What follows logically from such an observation is
to the settings they purport to describe, and itfidel ity to the settings of socIal that qualitative field research should pay careful attention to the collection
balance if only for the sake of completeness and and . analys�s of documentary realities. Such enquiry is not confined just to
research.
Organizations and other research settings have a variety of ways ?f the ll1spechon of documents themselves (important though a close scrutiny
representing themselves collectively both t� them selves and to othe�s. :t lS� must be). It must also incorporate a clear understanding of how documents
orary SOCIetIes are produced, circulated, read, stored and used for a wide variety of purposes.
therefore, imperative that our understandmg of contemp
57
PAUL ATKINSON AND AMANDA COFFEY " TEXTS
The production and consumption of documentary data has formed a part
of qualitative analyses of a range of settings. Well-cited examples include w ?rk
that has incorporated analyses of school reports (Woods, 1979), medIcal
records (Rees, 1981), classifications of causes of death (Prior, 1985) and health
visitors' case records (Dingwall, 1977). Indeed there are many research
questions and research settings that cannot be investigated ad�quately without
reference to the production and use of documentary materIals. It would be
fruitless to study the everyday work and occupational culture of a profession
such as actuaries without addressing the construction and interpretation
of artefacts such as the life-table (d. Prior and Bloor, 1993). Likewise, the
ethnographic study of accountants would be jejune without reference to
the professional use of accounts, book-keeping techniques and so on. Mor.e
generally, as Bloomfield and Vurdabakis (1994) point out, textual commUni­
cative practices are a vital way in which organizations constitute 'reality' and
the forms of knowledge appropriate to it.
In paying due attention to such materials, however, one must be quite
clear about what they can and cannot be used for. Documents are 'social facts',
in that they are produced, shared and used in socially organized ways. They
are not, however, transparent representations of organizational routines,
decision-making processes or professional diagnoses. They construct par­
ticular kinds of representations using their own conventions. Documentary
sources are not surrogates for other kinds of data. We cannot, for instance,
learn through written records alone how an organization actually operates
day by day. Equally, we cannot treat records - however 'official' - as firm
evidence of what they report. This observation has been made repeatedly
about data from official sources, such as statistics on crime, suicide, health,
death and educational outcomes (Cicourel and Kitsuse, 1963; Sudnow, 1968;
Atkinson, 1978; Roberts, 1990; Maguire, 1 994). This recognition or reservation
does not mean that we should ignore or downgrade documentary data.
On the contrary, our recognition of their existence as social facts (or construc­
tions) alerts us to the necessity to treat them very seriously indeed. We have
to approach documents for what they are and what they are used to accom­
plish. We should examine their place in organizational settings, the cult.ural
values attached to them, their distinctive types and forms. The analysls of
such evidence should therefore be an important part of ethnographic studies
of everyday organizational life and work. Of course documentary work
may be the main undertaking of qualitative research in its own right (Prior,
2003). In either event it is important to establish a methodological framework
for the analysis of documentary realities. In the remainder of this chapter we
outline a number of complementary strategies for approaching this kind of
qualitative data analysis. This is not intended as a .comprehensiv: revi:w
of all relevant empirical research or analytical strategles. Rather our mtentIOn
is to introduce some practical approaches to the systematic analysis of
documentary data and the contexts of their use.
It is important to recognize throughout our discussion that follows that
we are not - as are many of the social actors we observe - trying to use the
58
ANALYSI N G DOCUMENTARY REALITIES
documents to support or validate other data. It is tempting, when undertaking
ethnographic fieldwork or some similar piece of qualitative research, to
observational and oral data (such as may be derived from treat
interv
recorded interaction) as the primary data, and any documentary mater or iews
seconda ry. If used at all, then the latter are often drawn on to cross-check ials as
oral accounts, or to provide some kind of descriptive and historical contethe
Our vie� here, on the contrary, is that such attitudes to documentary xt.
.
data � re mapproprIate and unhelpful. We would urge that documentary
materIals should be regarded as data in their own right. They
a distinctively documentary version of social reality. They have often enshrine
conventions that inform their production and circulation. They are assoc their own
with distinct social occasions and organized activities. This does not meaniated
there is a documentary level of reality that is divorced froIl) other levels, suchthat
as the interactional order. Documents are used and exchanged as part of social
interac�ion, for instance. Nevertheless, it is vital to give docum entary data
due weIght and appropriate analytic attention. There are many ways
such documentary or textual data can be analysed and it is not our intent in which
to try to describe these in detail (see Silverman, 2001) . Rather, we introd ion
and exemplify a series of related themes and issues that can be brought to uce
on docume�t� ry sources. Our general perspective is informed by a broad bear
ethnographlc mterest, while our specific analytic approaches perhaps ly
more fron: a semiotic perspective. By that we mean an analytic perspectivee deriv
that exammes how documents can be examined as systems of convention
signs and modes of representation (d. Feldman, 1995) . Throu al
gh illustr ating
form
such an approach we consider how one needs to take account of the
of te�dual materials, the distinctive uses of language they may
.
relatIonshlps between texts and the conventions of genre. displa y, the
DOC U M ENTARY LAN G UAGE AN D FORM
Documentary reconstructions of social reality depends upon particular uses
of languag� . Ce�ta �n �ocument types constitute to use a literary analogy
genres, WIth dlStmctIve styles and conventions. These are often marked
by quite distinctive uses of linguistic registers: that is, the specialized use of
-
language associated with some particular domain of everyday life. Particular
? cc�pations often have distinctive registers, as do particular kinds of organ­
�zatIOn or cultural activity. One can often recognize what sort of document one
IS dealing with simply through a recognition of its distinctive use of language.
One can, for example, probably recognize the register of, say, a theatre review,
?r a wine appreciation, without seeing more than a random extract from
It. Each genre has its characteristic vocabulary, and reviews in general often
have characteristic form and tone as well. (The register of wine-talk is richly
faScinating in its own right, of course! See Lehrer, 1983.)
At a common-sense level we can recognize that official documents and
reports are often couched in language that differs from everyday language
59
 ANALYSING DOCUMENTARY REALITIES
COFF EY TEXTS
PAU L ATKI NSO N AND AMA NDA

sort of device that is u�ed to competition to gain the highest possible RAE grades. They devote considerable
use . Indeed, as we shall illustrate, this ise one of documentary representatIOn. It �ime and �ffort in long-term preparations for the RAE, and in the shorter term,
construct the distinctive and special mod nation of 'officialese'. or to �ssume m preparmg the actual documentary submissions.
is not necessary to endorse a glib condem or mislead through their specIal us�s The RAE is handled by the Higher Education Funding Council for England,
that bureaucracies deliberately confuse unhelpful to approa�h the analys�s on behalf of all the UK funding councils. Each submission has thus far followed
of written language. Indeed, it is usuallyally cal or evaluatIve stance. It IS � unifo�m format. For each assessment, the Council issues guidelines and
of documentary materials from an initi pt criti at the outset at leas� - a more mstructIons, and also issues standardized electronic forms that have to be used
undoubtedly more helpful to try to ado is to. p.ay �lose . attentIon to the t� prepare each assessment. The format of submissions has changed a little
interpretative standpoint. The initial task ted �s dIstmctIve kmds of �cro �ucts. SInce the process began, but there is a basic underlying structure to the returns.
question of how documents are construc to the textual orgamzatIon of !'his uniformity reflects a very significant thing about documentary realities
It is therefore appropriate to pay close attentIOn m the modern world. Documents are often used to create a certain kind
documentary sources. . . of predictability and uniformity out of the great variety of events and social
orde r to illus trate our ana lytic poin ts, and to loca te them withm a broader
In discussion with fragments of text arrangements. This is, after all, one of the most important features of the
cultural context we shall illustrate our of academic life - that is, the UK bureaucratic mode of social organization: persons and courses of action are
drawn from on� highly salient feature This is, in turn, an instance of how the reconstr� cted in terms of the categories and rules of the organization itself.
Research Assessment Exercise (RAE). in contemporary 'advanc�d� societies, !ransactlOns, problems, cases and the like can only be recorded and processed
'audit culture' has become prevalent beco me an especially vlsible asp�ct If the? can be ma �e t.o conform to standard formats . It is one of the key
not least in the public sector. Audit has s and pre�s�r�s, suc� as qua.hty functIOns of orgamzatIOnal documents that they do such work of standard­
of organizational life, as have related processe able that audIt m ItS vanouS gu�ses i�ation. v:'hen organizations generate documentary records, they transform
assurance mechanisms. Indeed, it is argu d reflexive mod ernity, or modermza­ dIVerse .Clrcum�tances an� people into documentary forms that can be pro­
is characteristic of what has been callemod ern social organization in. which cess� d In relatIvely predIctable and standardized ways. It thus becomes
tion. That is, a distinctive mode of r agenc�es are constramed to pOSSIble to apply various kinds of routines to processing them and recording
states, corporations, bureaucracies and othe ities and theIr consequences. These them, to generating statistics that compare and collate cases, problems
scrutinize and account for their own activ umentation and the textual artefacts and outcon:es. They can be used to do the sort of work that is currently popular
processes rely primarily on acts of doc e�sities �as not escaped these new among p.O�lCy makers: s�tting achievement targets and measuring outcomes.
that are generated. Academic life in univ mcreasmgly called on to account Such pohtlcal and orgamzational work is impossible without the construction
demands. On the contrary, academics tare have done. These accounts are of documentary facts and realities. Standardization and categorization through
for themselves' and to document whabodthey are charged with scrutiniz­ documentary types operates in many domains of modern life . Bowker a�d
used as eVidence by various external iesinthat academic depar�m�nts. O�e �f Star (1999) provide a particularly telling case study of how classifications
ing the 'quality' of research and teaching its i� the �AE. ThIS IS a penodic of . disea� e actively shape modern medical thought and practice (see also
the most high profile of these external aud ssment (whIch may or may not b.e Pn.or, thIS volume). Chaney (2002) discusses, from a very different starting
review of the research of 'units of asse emic disciplines �n all UK i�stl­ pomt, aspects of the documentary imperative in modern culture and its
coterminous with 'departments') in all acad n place at regular mtervals smce rationalization.
tutions of higher education. It has take
1985 . The most recent RAE results were pub lished in 2001 ; since then the whole . A great deal of work in modern bureaucracies and other complex organiza­
RAE process has been subject to a revi e-:v and pr�p � sed revi�i?n (Robe:ts, hons would be impossible without the creation of these kinds of documentary
mg exerCIse m the OffICl�l reflectlon reality . . Indeed,. the development of modern methods of accountancy,
2003) in an exercise that is itself a fascinat for each RAE, academIc staff and marketmg, tradmg and so on has depended completely on the construction
on documentary realities. In preparation ten submissions. �he.se are then sent of standardized categories and types - largely managed through the creation
university administrators prepare writeva e the submIssIOns and award ?f sta��ard doc.umentary forms. (The congruence of meaning of the term
to panels of discipline experts, who 4, luat and 5*. (�his distinctly o� d scale forms 1.S revealmg here - referring as it does to both types and the paper or
grades, ranging from 1 to 2, 3a and 3b, ng 5pas deciSIOns, compron lses and electromc :neans through which a typical case is constructed.) Documentary
is a very British phenomenon, reflecti goest up to 5!) The RAE ::IS a very reportage I� closely related to the existence of official classificatory systems,
additions: it is a seven-point scale that in it. It has profound consequences such as the mternational classification of diseases (see Bowker and Star (1999)
serious exercise for all those who take part ent university) is hel� . More for a detailed discussion of such classificatory systems). It is important to
for the esteem in which a given departmnt frac(and tio� of t�� state f� ndmg th�t reme�ber that such official categories and classificatory systems do not simply
fundamentally still, it drives an importa umversitles are mvolved m descnbe classes and systems - they are active in creating and shaping them.
is allocated to universities. Consequently, all
61
60
 ANALYSING DOCUMENTARY REALITIES
COFF EY
PAU L ATKI NSO N AND AMA NDA
�; TEXTS

At first glance Example 4.1 is a relatively straightforward table. While it

:
The author(s) o R AE submissions� for example, had restricted scope in termes
ten guidelines that were issued, ther
does not textually describe in detail the source or application of research funds,
of content and r As well as th writ that it is set out in a particular way, following some obvious stylistic conventions.
we� al�o the e� :tations and und erstandings about such documents, rea
t to bear. As drafts are comp�sed re�
It does not contain lengthy descriptive prose or explanation. It is tabulated in
aca emlCS and a�ministrators brough individuals con. tribute then sha
an ordered and structured manner. The RA4 statements will look similar
and com me nted upon internally' dsocult l as�ur�ptIOns. in terms of style and layout across all of the submissions. Like the layout, the
understandings of such backgroun entura sIOn there are a. number of sources of income detailed in the statement will be familiar to an academic
As art of an RAE unit of assessm i submIs . readership, and perhaps less understood by a lay readership. Moreover,
: !:
�� :� n b����:: �:����::' :s�:��� :��:�::;�::�e,a�env��Iron;i�men
:
t
; :�� :�: r; ! funding from the Office of Science and Technology (OST) and the research
councils (such as the Economic and Social Research Council) is held in par­
structur t and
dIs. �u:sIVe
' repo�t of for example the research ticularly high esteem, and this is perhaps implicit in their position within
activIty. Eac : �he�e can be comisn'sidered as representing a particularforkind
0 lity The sub sions for each unit of assessment the the table (which, incidentally, constituted for this particular submission the
of dO
: :o.a� . uk/ rae/ largest source of funding) . This positioning at the head of th� table hence serves
0 1 �� ::��::7r �ow �ublicly available (see http :/ /ww w.h em
� � x htm) For the purposes of illustration the examples used thIS chap.ter a further purpose in the construction of the research account.
E EducatIOn Presented with a file of statements like this one, and others relating to
��e :li take� from the Cardiff University submission to the RA staffing numbers, research studentships, students and so forth, the academic
Panel (20�1)� . RA4, which details research inc�me reader or panel member can read a representation of the unit of assessment.
Examp e . ' 1 IS a r�. production of form This sort of documentary reality might be of limited value to a lay reader.
over the reV iew penod. What "'!e are not . concerned with here is a detaIled The competent professional reader can, however, read into such documents
�na1�s�. s °f h w much research mco me IS clalm ' ed. by the table (though that
her we can thmk
. much more than the lay reader may be able to. The forms also enable all
IS no : de: that that may be a sign ifica nt fact or). Rat
knowledges .and assump­ submissions under each unit of assessment to be readily compared, albeit in
t a � in which this form reproduces sharedcase a particular, coded and highly stylized way. The RAE, in this way, involves
;
��:� �o;: articular kind of audience (in this academIc
peers and
the creation and use of particular kinds of documentary artefacts, as a way
administrators) . of constructing and representing academic departments (or parts thereof).
Moreover, this codification embodies within it assumptions of 'what counts'
as research activity or esteem, as prescribed by the funding councils (and
perhaps the academic community at large). Tabular summaries like this
Exa mpl e 4.1
can, therefore, provide standardized, shorthand ways in which the relative
RA4: Research income details
performance of different organizations can be compared . They thus inscribe
implicit assumptions about what is important about organizational life, and
) Partial 1 996-97 1 997-98 1 998-99 1 999- Partial what is worth recording. Moreover, such tables provide numerical summaries
Source of inco me (Net VAT 200 0-0 1
1 995-96
2000
of a wide range of activities. Obviously, there is much more to academic perfor­
mance than such summary statistics, but the use of such 'performance
al . 1 64 ,0 1 4 1 89,0 1 8 209,753 94, 2 1 0 92,1 1 4 23,459 indicators' is a very common feature of modern documentation and modern
OST research coun cils et
1 ,201 management. It is a representational strategy that is fundamental to the
rationalization of modern life (Scott, 1990).
4,767 1 4,075 1 2,663
1 , 389 2,635
UK- based charities

UK central gove rnme nt 1 22,038 1 44,81 2 25,805

Tabulated data are a particular form of documentary source. Other parts
54,95 5
bodies
74,4 47 39,7 43
of the RAE submissions by units of assessment relied upon prose text rather
UK industry, com merce
and
25,755 1 , 786
than figures. Prose accounts are equally conventional and deserve similar
33, 1 45
public corporations 1 ,7 1 8 2 1 ,786 34, 1 43
attention to matters of form. This can be illustrated by another example from
0 6 1 6,906 0 1 ,963 0 the RAE submission. Again we use the Cardiff education submission with­
EU gove rnme nt bod ies
0 0 1 1 ,668 0 0 out evaluative intent, but simply as an example of a document of a particular
0
EU other
4,54 7
kind, displaying characteristic features. Example 4.2 is an extract from form
8,90 4
Other overseas
0 2,98 4 1 , 1 67 0
RASa, the description of the structure, environment and staffing of the unit
2 1 ,01 5 3 1 ,535 1 7,926 47,022 1 4,87 2 1 0,885 of assessment. This extract is part of a broader statement, in which the claims
Other
that are made are justified and amplified.
63
62
 ANALYSING DOCUMENTARY REALITIES
PAUL ATKINSON AN D AMANDA COFFEY .� TEXTS

of learning taking place in a variety of contexts, and of interdisciplinarity are

'interpretations' of the educational research environment that particularly
Exam ple 4.2 suits the kind of educational research that is conducted at Cardiff. Similarly,
drawing attention to the view that educational research might have tenden­
RA5 a: Structure, environment and
staffing policy cies to be 'intellectually isolated and dominated by its teacher education
som e time adopted an expl ici analy tical : environment' is purposive given the fact that the Cardiff education unit
of assessment is not a centre for initial teacher training, unlike many of its
Educ ation al research at Cardiff has for h
a 'crad le-to -grav e' proc ess" wlt,hln whic
ers ectiv e that view s educ ation as � comparative units or schools. Beginning with these kinds of statements is
� ! I� see a�
very impo rtant , part. Lear ning
ch ling constHutes only one, albe it
socia l cont exts; in scho ols, colle ges and UnlVer ltles
,
� not accidental, therefore, but is designed to 'set the scene' in a very particular
takin g place in a variety of ,ity setti ngs
, way, The text implicitly invites the reader to find a point of contrast that can
" work -plac es and wide r com mun
00
certaln' Iy, but also withi n hom es ply withi n the be used to construct the distinctiveness and special quality of the Cardiff
'
� ?
are unde rs t d no t slm
The i m pacts of educ ation al chan ge
m itself , but also ?:
i n relat ion to her elem ents o,f th
s clal
, department.
,
al syste
,
educ ation We can also examine the kind of language that is use..d in the extract. It
ral Instit ution s,
' I
ets and politi cal and cultu
struc ture, suc h as fami'I'les, labo ur mark contains many words and phrases that will seem very familiar to an academic
educ ation al research wlthl a, strong socia
n
This approach necessitates situa ting ,
I IC fram ewo rk, whic h fosters the deve
, t'f' lopm ent of inter disci plina ry work , readership, For example, 'a strong social scientific framework', 'the develop­
sClen
ation al rese arch to be b ot h '
In t � il ec t u ally m ent of interdisciplinary work', 'capacity to deliver', 'research-active staff',
This counters any tend encie s for educ ,
antly teac her educ ation enVlr,onm ent. These, and expressions like them, are not necessarily part of everyday talk
�� �
isolated and dom inate d by its pred omin
ce In an
improvement of poll CY a ractl (even of academics) but they will be familiar to academics charged with
It also facil itates cont ribut ion to the
era 0f 'I ncre asing I Y 'J' o'i ned
' " u p' gove rnme nt and ,colla bora
, ,Initiatives between
tive authoring and reading such documents, They are among the linguistic
profession al disc ipline s, building blocks of this particular kind of academic (re)presentation, Indeed it
Scie nces has boosted capacity to deliv
er is possible to think of them as resembling verbal/textual formulae and their
The creation of the Scho ol of Soci al
The Scho ol bri ngs t geth ?
er almo st 1 00 deployment in documents can be likened to the composition of an oral epic.
this amb itiou s rese arch agen da, , �
socio logy , soci al o llcy, socia l work ,
, and It is well documented that bards can extemporize the composition of lengthy
rese arch -acti ve staff in educ ation ,
y distin ctive enVlr nmen � t for edu atlo� �
al oral verse, celebrating themes such as heroes or wars, They can ' compose'
crimi nolog y, It therefore provides a highl logis t,
researc h , U n der the leade rship of
the Direc tor, the distin guish ed socio line after line, usually in strict metre, on stock themes, peopled with stock
B non it has developed a miss
, rch
ion to produce Inter d"ISCIP I' Inary resea characters. It has been shown that this capacity for 'improvization' is possible
�: 7: �
H cont ribute s to
w h the retica lly i nform ed, methodol ogica
lly rigor ous and
rcher s
because the bards have at their disposal a repertoire of stock phrases for
ation al resea
the deve lopm ent of policy and profe
ssion al practice: Educ
rch agen da,
recurrent actors and actions, The oral bard, therefore, does not actually com­
this wide r resea
to make to pose verse from scratch, but rather puts it together from well-established
have a critic al intell ectu al cont ribut ion
�
whic h the r latio nship s between theo,
ry,
from a disc ipli nary back grou nd in
?
y and practl e are well velop ��
ed relat l � e patterns and pre-formed components. These are, in turn, familiar to the
methodology and appli catio n to polic fican t elem ent I n audience, who are able to draw upon the same stock of well-understood and
s, This will be a signi
to many othe r socia l scien ce disci pline , well-established phrases and literary conventions, In that sense, such a genre
ol s research strat egy,
the long -term deve lopm ent o f the Scho constitutes a kind of restricted code (Bernstein, 1981), with limited possibilities
of choice and combination in composition, Such an observation is equally
applicable to written language, as the RAE extract demonstrates,
then can we begin to make sense of suchhtextu al doc�mer:tary The author(s) of the Cardiff education RAE submission clearly had some
H rt IS wntte�, leeway in writing the account. (It is worth noting here that the submission is
ate0:"11' We �ight begin by noting the stylenarrin whic the repo
reader IS not attributed to named authors, but comes from 'the school', or indeed the
r :�y ways this follows a conventional ative form at. The
focu s of the university another feature of official documentation.) Despite this freedom,
:er:nded that it is educationesalfrom
research at Cardiff that is
what appea� to be gene
the
ral st � teme nts however, the scope and language of the document will have been restricted
iece , The extract then mov
;bout educational research, to mor e particular detaIl,s about the Card � ff SCh �� ! in some ways. There were written guidelines that indicated the kinds of things
that needed to be covered, Moreover, there are shared understandings and
enVIronment. The plece e1
of Social Sciences, its research philosophy and
a

towards the future. On , expectations about such documents and their use among academics and higher
a ain with a mor e general state men t and a look education administrators. As drafts are composed, read and commented upon
1 " however many of these general statemen ts are actually
c19oser. examina
' t'on
at Card
,
Iff ,
IS able to I ocate internally, so relevant individuals bring to bear these shared meanings and
d ' a canvas onto which educational research ve' process, cultural assumptions.
Ft:��� ;�� views, for example, of education being a 'crad le-to -gra
65
64
PAUL ATKINSON ANO AMANDA COFFEY w TEXTS
ANAL YSI N G DOCU MEN TARY
,
REAL ITIES

Most importantly, perhaps, we can now see that while this RAE document however tangentially or at one removed to oth
(or extract thereof) may be 'about' an academic school (�r 'uni.t �f assessment' also refe: to other doc�ments. This is esp er realities and domains. They
. ecia lly, tho
- an RAE category that could be analysed in its own nght), It IS not a trans­ of org�mzatIo nal sett�ngs and their systems of recough not exclusively, true
parent description . One certainly cannot take the doctlment and read of! from of documentary realIty must, therefore, look rd kee ping. The analysis
it a picture of the department and its academic/research pr?gr�mme m any a.sk how they are related . It is important to reco bey ond sep arate texts, and
simple sense . That is not because the author(s) set out to deceIve. m som� way. sI�ns and messages, documents make sense bec gniz e that , like any system of
The issue here is not about honesty, or even about accuracy, m any SImple WIth other documents. In that sense, therefore, aus e they hav e relationships
we can exa min e such
sense . It reflects the extent to which documentary realities constitute distincti.ve �nd their significance in just the same way we approach the signs of artefacts
levels of representation, with some degree of autonomy from other soclal Itself. language
constructions . Thus the RAE submission is a particular kmd of account of an Perhaps it is useful here to return
academic department, in just the same way as a finan�ial a�dit is a particul�r might be considered an example withtointhe theme of aud it of which the RAE
kind of account of an organization. The accountants audIt of a company s UK
the mechanics of audit, then it starts to become higher educa ;ion. If we consider
books and balance sheets is not intended to authenticate every record of of syst��atic relations between documents. One quite easy to grasp the point
every transaction, and the auditor is not fo:ever comparing the '�ooks' with an �ud1t IS that of the aud it trail. Traditionally of the root metaphors of
some independent level of reality. The audIt compares record.s .w1th record�, of flr�s and orga�izations, carrjed out by account (financially) defined audits
and checks them for features such as consistency. A key aud1tmg process IS on �h1S audIt. �raIl. At the begmm . . ng of ants, place great emphasis
the referencing of the audit file, a process whereby an auditor (usually a junior thei r trai ning, junior accountants
�re mstructed m how to carry out a detailed audit trail. This invo
accountant) is charged with the task of checking for consistency . an� order mg each document and statement pre ted lves retrac­
across the range of balances and documents presented to the audIt. LIke �he oth�r docume�ts cor:tain d in the auditsen in the 'company' accounts to
RAE, it too is conducted primarily within the domain of documentary realIty. audIt). There IS an ll1bm�lt assumptionfile (the preparation of papers for an
that
Such a conclusion does not mean that we cannot learn a great deal from such made to other documents. An auditor's task is reference can and should be
documentary sources. On the contrary, we have already sketcl:ed �ome ways relation�hips �nd intertextualities. 'Ticking' is to establish the extent of these
in which an understanding of the RAE document, and others lIke It, can start �o descnbe thIS process of retracing links anda folk term used by accountants
to give us valuable data about organizational life in academic in�titutions IS paralle�ed by 'b�shing', which refers to the acti establishing an audit trail. (It
and their characteristic cultures. We can, moreover, learn somethmg about of an audIt as conSIstent and fair .) on of signing off segments
the sort of work, and its associated skills, that goes into the creation and use A�ademic auditors foll ow similar procedure
of such an artefact. pOSSIble for an academic auditor to pick s. It therefore ought to be
What is at stake here is the construction of a distinctive documentary an item ,
decision, an� follow a 'paper trail' through the such as a transaction or a
. s, mmu app riate pro cedures of
reality. Our RAE document performs quite dist.in�tiv� work. It creates a very meetll1g tes, accounts, and so on, in order torop
. trace an orderly and
particular version of reality. It is, moreover, a d1stmctIvely documentary one . p rope:ly �uthonzed procedure which is in turn correctly acco
It draws on the genre of other documents, and their characteristic lang�age. mvestIgatIve procedures are predicated on the unted for . Such
It is constructed so that other kinds of documentary work can be accomphshed should be regular, identifiable relation assumption that there are and
. s�lp . are bas ships between documenta records.
with it . Such a self-assessment and self-report are eventually used as evidence These relahon
. � ed on elementary - but significant _ ry
They mclud� pnnc1ples of sequence and hier principles.
archy. The se in turn are part of
by a panel who in turn generate yet another docum�ntary reality - thei� �verall
report - and perhaps more i�po:tant tha� anythmg else to the part1C1? ant� the constItu. tIve .
machll1ery whereby anizations produce and
- the final RAE grading. ThIS bnngs us, mdeed, to our next analytIc. Issue. themselves . From a general analytic perorg reproduce
spective, therefore, we can see that the
documents do not exist in isolation. Documentary realitydepends on system­ realm of do�umenta�y reality does not rely
atic relationships between documents. Analysis must take accoun: of s�ch and reflectmg a SOCIal reahty. Rather, weon particular documents mirroring
relationships . We now turn, therefore, to a discussion of the relatIonshIps domain of documentary reality, in which doc can think of a semi-autonomous
between documents, or the intertextuality of texts. implicitly) to other documents. uments reflect and refer (often
Colleag es i� acade�ic life will be perfectly
of the aud�It pnn . em
CIpl
familiar with the application
contemporary universities . All univers
I NTERTEXTUALITY �epartments find themselves subject to vari ities and
III recent yea rs m . the UK ous forms of audit. For example,
academic departments found themselves sub
Documents do not stand alone. They do not construct systems or domains to assess� ent by �he Quality Ass ject
of documentary reality as individual, separate activities. Documents refer - of Teachmg QualIty Assessment ura nce Agency (QAA) through the process
(TQA) . Universities continue to be subject
66
67
 ANALYSI N G DOCUMENTARY REALITIES
NDA COF FEY " TEXTS
PAU L ATK INSO N AND AMA

" IOn al audIt' by the QAA ' This roc, ess of TQA,obs for example, invol�ed sequences of decisions and their consequences, distributed regularly over
to instltut
a certa in amount of direct eviden �e ga� rmg' by the ervation of alteaofchm g time, and reported in uniform formats, We could thus examine how docu­
mentary realities have temporal dimensions built into them, Note that this
gs with students an teac��rs ' tion However a great de the
and meetin of do�umentary material s , is an organizational or documentary time, it does not describe the passage
inspectIOn t00k place througdh tthhem
, e examm a
. seIves producing self-assessmentvid d ocu- of time as experienced as an everyday phenomenon by the individual actors
Academic departments foun
" l1ar a process from the l'AE) as well ash as module pro ing concerned. In another sense, documentary sources suppress time, by lifting
ments �not too d'l �slm
" � , ,

ts of docum entary evidence (suc events out of the flow of lived experience, and recording them in the de­
inspectlon teams, wIth vast amoun s of me gs, examp1es 0f students'
, work' contextualized language and formats of official records, Intertextuality thus
outlines and read'mg rIStS' minute etm
was to evaluate the extent to w hich alerts us to the fact that organizational and official documents are part of wider
�,
;
and so on), A ,task �f th�insr�� ��ht�� was evident through the supporting systems of distribution and exchange, Official documents in particular
a department s clalms a ou i circulate (though often in restricted social spheres) through social networks
documentation " use documents which in turn help to identify and d elineate divisions of labour and official
When aca dem Ic , au�ltor ' s, and, similar external people min
education' they exa e a range ofs, positions. One important analytic theme here is the o@servation that the
to scrutinize orgam� atl,ons f �gher � systematic relationships between documents actively construct the rationality
sources, As we ha�e mdlcate��izatrona undertake an audit trail. In simple term
l decision, an innovation or a probleme and organization that they purport transparently to record.
that means followmg an orga
' u ence. of document-s, Such' a trailermig ht, for instance, examin Documents can circulate and be exchanged partly because they are used
throug h a seq
the minutes of departmental meett'mg s ln ord to trace the progress 0ords f an to de-contextualize events, We transform things by incorporating them into
ation al rec texts, By writing something in a documentary format, we translate them from
item from o�e meet'mg t� the nex " " and so on ' Such otrganiz
a �n, we can note tha such document s have the specific and the local, and make of them 'facts' and 'records' which take
have distinctIVe charactenstIcs, Ag a sO have particular fun ctio ns, Min utes on an independent existence, Some texts become 'official', and can become
specific,, stylized formats, Theyreco r d ,everything that was said rd wehin and don 'proof' of events and identities, This point is made in relation to the production
' stance, do not
of meetmgs, for m eClse1 are not intended to reco
a m eeting , Indeed, in a sense thewYhPatr was dYecz'ded In a sense, they constitute
at of scientific facts and findings by Latour and Woolgar (1986), who write about
the production of scientific papers, and suggest that they achieve an
was actually �al'd , They r ecord11e,n ged and corrected ' with the agreement of the independence of their original site of production - the research group, the
what was dec Ide d, U nle ss ch a
cedence over rnembers' own recollec-r laboratory - and take on an independent existence, The accountants' audit of
members, the wn'tten record takesu pre ments are written in order to refe a business organization takes on a similar existence, The audit report becomes
tions and inte�tions , Moreover�: �hhedo��e co�structed and read precisely
to other, eqUIvalent doc�me n " of l' �erlinked documents, If we pursue the
the documentary reality, superseding other files, records and memories,
Similar observations could be made about the RAE submissions where units
as part of a documentary omaca am �
hypoth etlca' 1 ex amp I e of our d.emlC aud't1 , we can see that an aud it , trail of assessments become 'real' and the documentary reality that is presented
' k up 0n docum.ents suc11 as minus tes of staff-st ude nt mee tmgs, supersedes other accounts,
wo uld pIC , rou of them), We
meetings of the a�a,de�mc st�ff s(orwasubs �e [ave already referred to the notion
can ana lyse such Documents are also written and read with reference to other occasions
of use, They can be referred to in order to warrant or challenge subsequent
documentary �eahh�s var� ouderiv:d from contemporary literary criticism, actions and decisions - possibly long after they were first constructed, They
of intertextualztY' , T�11S term IS f t that fact that literary texts (such as novels)
m

can also inscribe positions of hierarchy, Documents report discussions,

in which context l,t IS used to re �:y°do not refer J'ust to a fictional world, Rather, decisions and events to people or bodies that are superior to the originators,
are not free standmg, , and tha' t t to other texts, They l'nclude other texts of the
they refer, howe�r l;P;l' Cltly extual product (such as journalism, biography,
The right to construct a document, to challenge it, to receive it and act on it
(or not) is part of the formal division of labour within many social settings,
same, genre, or ot er m s Of ; lyse texts in terms of these intertextual One cannot 'read off' such organizational realities from documentary sources,
mov:es) , W e can , therefore ana and the attempt to do so would be based on a fundamental misunderstanding,
,
relatIOnshIp tracmg the dimensrea
s, ions of similarity and difference, rk settmg , ,
0 org ani zati ona l or wo Documentary realities, based on complex inter-linkages between documents,
create their own versions of hierarchy and legitimate authority, Indeed, the
' g the documentary l Ies
l't' f a n
In analysm e hOW
therefore, can explore the 1'nte
we rtextual relationships, We can examin
d between texts and thu s how they con struct issue of authority raises for us the closely related issue of authorship and
conventional formats are shareW note ho� they are link ed as seri es or readership, to which we now turn,
a unl'form, bureaucratic style, ' es 0can e11' ngs refer to prev ious min utes and
sequences 0f document, s" Mmuteutefsme different meetin gs wil l loo k rem ar k-
things like 'matters ansl�g' , Minguageofan d tone , They thus construct rational
ably similar in constructIOn, lan
69
68
PAUL ATKINSON AND AMANDA COFFEY ,' TEXTS
ANA LYS ING DOC UM ENT
ARY REA LI IES �
�xpIca
.restiosinnsg op inions, be
of authorshiplie' Ofsr an . pect texts for
AUTHORS H I P A N D READERSHIP
Ind d so on' W. e can therefore Ins
't1s a b
The kinds of documents we have been discussing may have identifiable, how they claim whatever author't sen ce' In that. sense' t00, We can look for
��';!u
individual authors or they may be anonymous, even collective, products. this point we can return to the RlI &,b� thI at�nb.uted to them. To illustrate
Equally, they may be addressed to specific individuals or they may appear to prepared by the Education RAE ' e1 xa S tIme to the overview report
address an impersonal world at large. In any event it is important to address provided a general overview rePo!t�: l
mp!e 3) Each subject panel
4.
authorship (actual or implied) and readership (actual or implied) if one is to '
as well as specific fee' db. ack for eac umt of assess the sub mIssIOns that it had received
. .

understand the overall system of production, exchange and consumption of 1

ment. '
documentary materials. Documents, like all texts and utterances, are 'recipient
designed'. That is, they reflect implicit assumptions about who will be the
'hearer' or reader. The implied reader does not have to be an actual individual Exa mple 4.3
person. In many ways the implied recipient corresponds to what George
Herbert Mead referred to as the 'generalized other'. It is a basic tenet of Education Panel RA E 200
1 : Overvie w Report
interactionist social analysis that social actors monitor and shape their actions
in the light of generalized others' imputed responses and evaluations. So when Th e str ong est research
sub mitted to the E dU? '
atl �n Pa nel was cha rac
we create a document, we do so in the light of the kind of readership we are by its gro und ing in hig
h qu ality me th 0 d 0 Iogl terised
expecting or writing for. theoretical concerns and
its read Ine
es,
' ss to eng age in broad
Its u nde rpi nni ng by sou
. nd
In the case of the RAE submissions, for example, the readers are both a b out enq uiry in the fiel
d. Thi s was most eVl' den' SOC ia
' I sCi enc e debates

;
, sio
specific and generalized. The membership of the panels of experts is known a breadth and depth of ,
me tho d o I o lca l exp ert
t wh ere sub mis
,lse
(bo
ns d isp layed
th � ual itative and qua n,·
to those constructing the submission, as this is public knowledge. But the tita tive ), a wil l ing nes s ,
to asso cia e Wl, th bro
ade r SOC ial sci enc e or
bureaucratic formats and other conventions governing the preparation of dis cip linary fiel ds and
of exc elle nt research .
an eng age me nt Wit, h com
�
mu n !'t'les 0f users on the
oth er

the documents preclude a highly personalized appeal to particular readers. :m ine nt aca dem ic jou rna ls,
Th ese hi g h rat e su bm ' ,
iss ion s typ ic� l I� pu blis
basis

So even if the head of an academic department knows the chair of the expert Inte rac tion wit h the edu
bO O k cha p ers and out hed in
lets proVid ing sig n ific ant
panel of reviewers, he or she cannot construct the submission as a personal work was pub lish ed in
cat i on com m u n ,ity mo
research reports t 0 fun
re gen er� II y. S om e very
g ood
appeal to that specific individual. However much authors might dream of it, not always easily access d 'Ing bo die s alth oug h
ible to th e gen eral pub lic. this was
they cannot actually write a personal appeal to the panel members. It has to
be couched in the appropriate register, suited to the institutional demands and
expectations of the exercise itself. Indeed, it is part of the skill in constructing It is notic le that these statements are made .
such documents in administrative contexts that one should be able to use the to the persoeab ency that was a t ' 11 sponsiWIblethofor ut explicit reference
right kinds of phrases, deploy the right kinds of arguments, and generally t�� report. (Thnael enagtire ent is �e��i� 0r;name� authorshp�Ip,ttinangdtog ether
convey the right sort of tone. In audit exercises like the RAE, this includes the tlfIed solely with 'Educdoaticum is ide
use of distinctive terms and ideas that are intended to reflect the coherence of but not attributed in this repPao�;I') 'ThThee ;epemortb�rslup of the panel is known­n,
on
a department's research strategic thinking and the cogency of its research statements that have a similar for�1 to those rep IS made up of a number of
plans. While the self-assessment part of the document is intended to convey paragraphs / comments are less ? wm. g than roduced above, though some
a picture of the research activities of the academic department, in an important identifiable with an authorial 'vo�ce� as such. these. But the report is not
sense it creates the reality it reports. This report is normally not a highly seemed to us . . . ' or 'we felt . ThIS. lack We do not rea 1 d ph ras es
individual picture, but a highly predictable version, suited to the intentions c hara.cteristic of ma' ny 'off'ICla. I' 'd" ocuments Inof a pers ona au. th.o r IS. entlikiree ly'It
and the readership created by the RAE itself. keep In mind a distin
cti on tha t is ' thOIS conte.x� �t IS Important to
tak
.

en
between the author as a person ( m d or , rary cntIClsm: the contrast
,

While it is self-evident that a person or a group of people must actually f rom lIte
write/author documents (since they do not write themselves) that does not lvatter expression refers. to the te��ua� p :esenot) nce
and the 'implied author' . The
always imply a social recognition of 'authorship'. Indeed, it is part of the oice in the text. For InSt ance a paper In � sC (or absence) of an authorial
facticity of many official and organizational documents that they are not b� written in an impersonal m;nn lentl'flc Jou
' rnal will no
WIth no personal implied author, even t 10ugh ces such as the paSSI. Ve rm
. ally
'

identifiably the work of an individual author. Their very anonymity is part er (us
1 ing eVl vo ice ),
of the official production of documentary reality. There may be an implied of ames, may be
cre dit ed as its 'authors'. an a u thor ' s nam e, or a coI Iec tio n
'ownership' of a document - such as the originating administrator or depart­ �he absence of an implied autho
ment - but official materials do not normally have visible human agencies for the con struction of authoritativer, is'0ffonICla� �,horeto,nc' al d �vlc, e that is available
factual accounts. It implies
70

71
PAUL ATKINSON AND AMANDA COffEY TEXTS
T
I
ANA LYS I N G DOC UME NTA
RY flEA Ll II:S

a reality that exists independently of any individual observer, interpreter or CO NC LUS ION
writer. That device is not sufficient in itself to guarantee such status, which
also rests on the organizational, professional or bureaucratic contexts in which In the course of this brief chapter we have not tried to proVId. e
documents are produced and used. It should also be noted at this point that account of how to ana lys e doc u . ents and ot�er kinds of texa com prehensive
the overview report we have just quoted is a response to the RAE submission Rather, we have tried to indicat � a P1e, J t �ome of the tual materials.
we cited earlier in this chapter (and to other submissions to the Education documentary reality construct��n �:� � a �oac s ays in which
. The km:-v ds of thin
Panel). This sort of dialogue between the documents is a particular kind of we have discussed could well b d C ! hed P
intertextual relationship. Equally, and probably more fre����tl� ��e;i�����c�ses in thei\own right.gs
Documentary reality construction also involves implied readers. Any literate ethnograp. hic examination of orgamza ' . ' t·'lOnaI settinorm gs,'
part a broade
0
war1( practrc. es,r
person may actually pick up and read a document, but texts are often aimed professIonal cultures and everyday life.
at particular classes of reader, and it may be that only a restricted readership, Here, by way of conclusion, we recapitulate so� of the key
with specific competencies, will be able fully to decode them. This is especially points we have intr odu ced alo n the w a .rst, t �IS Impo
. methodological
true of organizational documents, where an understanding of the organization documentary reality does not �onsist Of � s �lph. ons f the to realize that
Y rtant
and its working assumptions may be a prerequisite to a thoroughly competent that can be used directly as evidence abO t. It.� �ne certaIn � ly cansocnotial world
reading. This observation reflects the fact that no text, whether literary or that documentary accounts are ' accura te� portrayals in th assum
official, can determine or constrain precisely how it shall be read. Reading is they constructtheirown kinds of reality It IS,· there.fore Imp . at s' ense. Rathere,
an activity, not the passive receipt of information. The reader brings to the text them as texts. Texts are constructed
.

. accordmg
c .

. to conven , . orta. nt. t approach

0
his or her stock of cultural knowledge, a knowledge (or ignorance) of similar selves part of a d tions that are them�
c

texts, and his or her unique biography. -I en

account is true, or :���:���7a:����. � s valId. , vldence abo
/c e, ra the �· than ask whether an
The examples we have already given can be understood quite transparently setting, it is more fruitful to ask our:el�es questIo: ns abo ut a research
from this perspective. At one level, any reader of English can make sense of function of texts themselves. ut the form and
them. On the other hand, it takes somebody who knows about research in Consequently,
higher education, and academic life more generally, to grasp much more of We have indicated,onevercany brieexamine texts for s�me f thelr. formal properties.
0

the significance of the phrases that are contained in the documents of the RAE. language used in such texts Mf!y,r how er,
one mIght look at the characteristic
e ca
Moreover, in the examples we have cited, some knowledge of social science rhetorica l features. SOciolo�ist� :�� �l � look at them in terms of their
0
t ers ave come to recognize with
in general, and educational research in particular, would make the texts increasing force that much can be lear e from
perhaps more meaningful. Such people are among the implied readers. Indeed Rhetoric is not t about the or�a en�atlO � n the diScipline of rhetoric.
the members of the Cardiff School of Social Sciences and those specifi­ speakers like poljusiticians Rheto s,� ndamentof�yspe, about
. eches, or the effects of
cally included in the education submission are also able to bring to bear their and written) ade th�ir rea;��; an� hearers. e all use how texts (spoken
personal knowledge and commitments, and will read the texts differently from in order to getpersuparticular pOI.Ilt . f VIew . across to others . rhe torical devices
readers who are not personally involved. ventlOns that area Wid
. 0
ely shared WIthOIn our cu1ture. We have toudcherawd ononSom W e con-
In some contexts, it takes a highly socialized member of a subculture (such of the ways in which d ocu
as a professional group) to make any sense at all of a text. The case of the a perspective. mentary sources can be examined from suche
medical record is a classic one that a number of sociologists have focused on. It is important to think about d . atIon . to their production
They all conclude that the medical record, or case note, is a partial and cryptic (authorship) and their consumptio:���;ntrs�IIp �n r�
text. Competent medical practitioners can make sense of their own or others' textual terms these are not cotermInO . US wIt� h the,parutfICU one should note that in
fragmentary notes because they bring to bear a wealth of context-specific, often actors who write and read W e d to ay c ose a tenlar In· d·IVld. ual social
tacit, knowledge about medical history taking, clinical examination, inter­ readers, and to lied · cla�: �f aut�ors�lp. ! � tion to the implied
pretation of laboratory test results, diagnosis and treatment. The briefest of interesting whentheweimp are exa min in g h W tex t Imp
��lS bec?mes particularly
lIcIt ly claIms a special kind
notes, often in the form of abbreviations, may have a wealth of routine medical of status -as . bJ?ectIve or sClen
tive with thatfacoftua l, authorI·tatlVe 0
�

work and interpretation read into it. ' · tifIC. L·m king this perspec�
rhetoric then, we can ask ourselves what claims a text see
to inscribe and wh �� ' . ms
claim Th: ';ame w� ��ec�srue are brought to bear in order to enter that implied
- though the rhetori. cal devices would differ
- if the document In. questIo. n had
a complaint, a confession or a persona d·ff tf
a� r:::is��1���� (such as constructing
72
73
 I
PAUL ATKINSON ANO AMANDA COFFEY TEXTS ANALYS ING DOCUM ENTARY REALITI ES

. another analytic perspective from literar y and

We have also . d�awn on sizin tuality . d that texts
REFERENCES

rhetorical analysIs m emph a g intertex

. We have stresse
ntial value is often
do not refer transparently to thcratl e s�cIa1 worId . Their refere Atkinson, J.M. (1978) Discovering Suicide: Studies in the Social Organization

to other. texts. In literate, bu�eau0f texZtsedansdettdm. .o��inm�nts

articular one may identify
of Sudden
Death. London : Macmillan.

a seml-autonomous domamrk of cross referencing, and that ;efer primarily to Bernstein, B. (1981) Codes, modalities and the process of cultural reproduction:
a model,
oneanother. A dense netwo . . - . shared textual formats, Language in Society, 10 (3): 327-63.
. t
create� a P?werful verSlOn :�i:��:��:�e�' we have implicitly drawn on the
Bloomfield, B.P. and Vurdabakis, T. (1994) Re-presenting technology: IT consult
reports as textual reality constructions, Sociology, 28 (2): 455--78.
ancy
In smglmg out �hese Pa the only. viewpoint from Bowker, G.c. and Star, S. L. (1999) Sorting Things Out: Classification and
1'deas and perspectlves 0 f 1't erary theory . This is not
1
its Consequences.
other hapters
c
Cambridge, MA: MIT Press.

.m th'IS tovoexami
which ne textual data. .YOU W�'11;�e are the best dletm alone
f d others outline
only Chaney, D. (2002) Cultural Change and Everyday Life. London: Palgrave.
1ume . We do. not cIaIm th010a ical analysis. It is, none the 1the
t
ess, our
Cicourel, A and Kitsuse, J. (963) The Educational Decision-Makers. New York:
- ways of approach�ng suche �OCI
_

' Merrill.
Bobbs­

contention that the Ideas 7(S fora��e0 examination of documents and similator
� tlined can be used and develo ped Dingwall, R. (977) The Social Organization of Health Visitor Training. London
: Croom
provide analytic framewor Helm.
cultural artefacts. Feldman, M.S. (1995) Strategies for Interpreting Qualitative Data. Thousand
Sage.
Oaks, CA:
Latour, B. and Woolgar, S. (1986) Laboratory Life. Princeton, NJ: Princet
on University
Press.
N OTE Lehrer, A (1983) Wine and Conversation. Bloomington: Indiana University
Press.
Maguire, M. (1994) Crime statistics, patterns and trends, in M. MagUir
1 . In the 2001 RAE, Cardiff Uni:ersity was one 0f two submISSIO
" ns to the Education e, R. Morgan
and R. Reiner (eds), Oxford Handbook of Criminology. Oxford: Oxford Univer
sity
Panel to be awarded grade 5 '. Press.
Prior, L. (1985) Making sense of mortality, Sociology of Health and Illness,
7 (2): 167-90.
Prior, L. (2003) Using Documents in Social Research. London: Sage.
Prior, L. and Bloor, M. (1993) Why people die: social representations of
death and its
causes, Science as Culture, 3 (3): 346-74.
Rees, C. (1981) Records and hospital routine, in P. Atkinson and C. Heath
(eds), Medical
Work: Realities and Routines. Farnborough: Gower.
Roberts, G. (2003) Review of Research Assessment. Bristol: Higher Educat
ion Funding
Council for England.
Roberts, H. (1990) Women's Health Counts. London: Routledge .
Scott, J. (1990) A Matter of Record. Cambridge: Polity Press.
Silverman, D. (2001) Interpreting Qualitative Data, Second Edition. London
: Sage.
Sudnow, D. (1968) Passing On. Englewood Cliffs, NJ: Prentice Hall.
Woods, P. (1979) The Divide School. London: Routledge & Kegan Paul.

74 75
 D O I N G THINGS WITH DOCUME�TS

almost exclusively as containers of content. Document content is important

and .there ar� numerous routes by means of which such content may be
studIed. "':'e should not, however, let the presence of content bedazzle us to
the exclUSIOn of, ot�er qualitie�. Above all we should recognize the quality
wi th of documents. as thmgs - as thmgs that can be produced and manipulated
used or con �med, an� � s things that can act back on their creators ver;
Lindsay Prior
muc.h as Dr?�<rankenstem s monster sought to act back on his creator, Indeed,
one lx:terestmg feature of documents in action is their tendency to exhibit what
we mIght call such 'monster-like' qualities.
In ,:hat follows, and for the sake of examples, I intend to emphasize three
speCIfI. � f�atures of document� as. a field for social research. They concern,
my pen: I write and it is done.
respectively, how .doc�mentat�on IS produced in socially organized contexts;
[Hjere I sit and govern [Scotland] with
es 1 (England»1
how docurr:entatIOn 1� used m everyday organizationaJ action; and how
(King James VI (Scotland) and Jam documentatIOn enters mto the manufacture of self and identity.
This is an instrucumtive quotation. It demtheonsegotratcenes,tricaboJamve alt how peoplet do
things with doc theent S. In the case of er and p n. �hese day es, it is clear tha he
nology of pap mumca � tIon I. addsltIo . �e woul� r:'0re AS PRODUCT
did things with electech tronic modes of com X: n to hIS mky Peopl� thin,k wi�h things as :v�ll as with words. How they arrange and
likely have used ial rese ers we could of c?urs e deC Ide to c ncentrate on
�On. orgamzec thmgs m the world IS Important. This, not least because - as was
quill. And as soc ures arch associated with themeaKmning gJam es quo tatI We could demonstrated during the 1960s by a range of French intellectuals such as
any number of feat ts of the man, and theLondon tothat J am es atta che d to. the Barthes (1985), Foucault (1970) and Levi-Strauss (1969) - the organization of
focus on the though t he dispatched from Edm bur gh. Or v e
: ludml?ht art�facts provides insight into the most fundamental aspects of human culture.
words on pages tha meaning of thosetes) who rea d the wo rds (mc mg
choose to focus onl asthe long-dead subordaninainterest . Or, as I hav e sug ges ted else ­ In lIterate ultures, of course, the organization of things is anchored in writing
as well as m� three-dimensional space. As a result the social researcher of such
ourselves, as wel 3), we ld eschew in mea nin g alto geth er and
where (Prior, 200 how thecouwor t contained them)erwer e cultures has � wealth of data available for analysis that is simply not open to
focus, instead, onof social interaction. In this chaptetha
ds (and doc um ent s
r it is t� e latt er rath t han
.�le, researc�ers m cultures where writing is absent though access to other
symbohc systems may well be available.
used in episodes that will be adopted. As such I sha � l aIm , w her e pos sI
. rpretatIOn For a panoramic view of the ways in which people arrange things in the
the former strategycern with document content, meanmg and mte world there �re few better objects for study than encyclopaedias. Nowadays
to side -ste p a con of text in a networ k of action.
on the position put
in favour of a focuens doc ration in �chemest encyclopaedlas are common enough items accepting that the weighty tomes
Normally, wh y um ents are forward sforof whside con
at the y contam. Tha of the past have been replaced by electronic rather than written media. Yet
of social research thecipaare approached in term document as a th world has not always been brimming with encyclopaedias. In fact, it seems
is, the focus is prin lly on the languaitgeisemb qui
odi
te
ed
clea
in
r
the
tha t each and every falT� to. assert that a set of volumes that claim to provide a comprehensive
medium of thoughinto t and expression. Yet tion. First, and astion with overVlew of the world and the contents within, is essentially a feature of
document enters manhum an projects in a duaaslarela receptacle (of instru�held ass, modern Western culture.
james' s royal comtracts,ds,wisthe y enter the field Second they enter the lo g There is probably little point in entering into debates about the first instance
obligations, conown right.hes , reports, etc) . do um, ent s . have eff�cts �Il­ of any one thin.g. Yet, it seems fair to suggest that one of the earliest of
agents in theiran creators areInddea eed, as ager:ts (WIl� ls prOVId e a readIly aVa all enc�clopaedlas was that published in Paris during the period 1751-72.
after their hum such effects) . Andd asandagebun ed always open to !he edItors f the vast work sought, above all, to provide a map of available
able example of others: as allies, as resourcnts,es for doc um ent s ar .
further actIOn, as eopp
e
onents kno,:ledge?and of all the byways within it. In short, to impose an encyclo­
manipulation by or suppressed. (We should not forg tl:at peopl. burnl role and pae�lc order on the world. In so doing they used a number of simple ordering
to be destroyed,as well as read them.) It is the examm�t aho of thIS dua �evlces., Fo�emost among the�e was the use of an alphabetic sequence - so
ban documentsintellectual backbone of the current chapter.n he entnes m the encyclopaedIa were arranged in dictionary order (Indeed
�n. t�e on� set of volumes v:re see the.er:'ergence of two rather impress'ive form�
that forms the stated, in so far as documents have been dealt wit� as a f hterahty - encyclopaedIas and dIctIonaries.) Alphabetic sequence, was not,
As I have just . been consldered
resource for the social scientific researcher they have hItherto 77
LINDSAY PRIOR TEXTS
D O I N G THIN GS WITH DOCU MEN
ST
however, the only ordering device that was used by the encyclopaedia's iS i is n Fo exam
editors. They also used an ordering metaphor to organize knowledge. That ����. �ey el?t�. er �ave rouple,ndethed cap
fungi either smell distinctly of fish or they do
s or they do not. This manner of eratin
metaphor was of a tree. The image of a tree of knowledge has, as Darnton �� b .nary pairs - sometimes referred 0
'

(1984) points out, a much older pedigree than that of the Encyclopedie. It Ica;IOn - was regar�ed by Levi-Straustos (196 as an Aristotelian system !t classi�
. rauss tried to decode 9) as fundam ental to all. human
is also of some interest to note that the only image of evolution provided in cultures. Ind eed, LeVI-St my th s an d totems In terms
Darwin's Origin of Species is of a tree, and that the image of a branching tree "
of such lOgIC' , pOIntIng ?ut how we should concentrate on how what is said is
continues to play an important symbolic role in Western science generally (see ordered, rat}ler than WIth what it 'means'.
Gould, 1989). In the case of Diderot and d'Alembert (the editors of the Paris Now the content of any classificatory system is important as is
Encyclopaedia) there were really three trees - named, respectively, Memory, Conse�uently, the ord er on the pag e requ ires ana lysi this one.
be ea Ily overlooked that order on the page is invariably tied into t can s. How eve r wha
_

Reason and Imagination. Within each tree there were many branches. So,
for example, the tree of Memory had branches leading to 'nature' and its of SOCI" � al order and It. IS.isthe con nect ion betw een the two that de ma n d
forms
many divisions. Within that there were divisions relating to the 'uniformity f19at IOn . For e amp1e, the use and
. � man ipu latio n of s '
mv es-
of nature', and the 'irregularities of nature'. Among the latter, we find, in turn, as . s, InvarIably serves to mark out the boundarIe' s f' SOCI.a 1 gro s,
tool clas sific atorY s Y stem
0
articles on monstrous vegetables, unusual meteors, wonders of the earth, (Kea ng an Cambro 2000), and in that sense they often fun upin
and so forth. boun� ary b�Jects (Star sio,
0 and Grie sem er ction �:
Monsters are not to be laughed at, and dividing or segmenting the world In this light wer . 1989) '
ciose.7' It. . wo�'ld qUIce. wklye tbecoexamemInevid
0
'
e the fungI. encyclopaedia a little more
into the routine and the monstrous is a serious matter. Above all, it is a form
of exercise that provides telling insight into how societies think (see, for c aSSI lcatIOn IS embedded marks out aent cert
that the frame within which the
example, Douglas, 1966; Ritvo, 1997). Indeed, one might say that the study ency
. th�clop aedia aimed at 'amateurs' (p . 7) dain eSIg
'
kind of user. In fac t, 1' t IS' an
ne d for use by the amateur
of taxonomic structure (or forms of classification) is key to the study of 111 f� eId , rather than, say, bench scie
ntis
outse , It prOVIdes the reader with a specific identity labo
. ts in the ratory. So, from
any culture precisely because it makes 'thought' visible. That is, in so far as
it can reveal a map of concepts and of the links and associations that are for reading (and dointhe g).
made between concepts. With that point in mind it is easy to understand
why the authors of the Paris Encyclopaedia often referred to their work as
a 'Mappemonde'. Yet, whenever things are divided, much more is implicated
than human thought and culture. For, associated with each and every classi­ Key C Basi diom ycot ina

ficatory system is a set of practices. And it is truly in the links between human
(Gill- bear ing and pore-bea
ring forms)

practices and forms of taxonomy that a space for social research opens up. 1a Hym enium conS isting of

In fact, if we re-conceptualize Diderot's Encyclopedie as an information

gills
1 b Hym eniu m cons isting
of tu bes open ing by pore

storage and retrieval system - rather than as a mere dictionary we will be

s

led to pose questions not merely about the content of the system, but also
ones about how the system was accessed and used and modified and
challenged. That is, how the system as a technology - is nested within a 24a Cap cam panu late or
coni cal
web of activities (Bijker et al. 1987). In recent decades, investigations into the 24b Cap not cam panu late
or coni cal

nature of information storage systems and their relationship to organizational 25a Cap grea sy or visci d,
brigh tly colo ured

life have emerged as a distinct subject area in itself (see, for example, Berg, 25b Cap not with this com Hygrocybe
bination

1997; Bowker and Star, 1999; Keating and Cambrosio, 2000). And as an
26a Sme lling disti nctly of
fish
26b Not sme lling of fish
introduction to just a few of the issues contained within this newly emergent
field it would, perhaps, be useful for us to consider yet another encyclopaedia.
Consider Figure 5.1. It is an extract from a table concerned with the manner
in which we classify what, in England, are often referred to as 'toadstools'. 30a On wood
Micr omp hale
The table is to be found in a modern encyclopaedia (Jordan, 1995). It is an
30b On cones
Baeospora, Stro bilur us

encyclopaedia of fungi, and it is designed to enable users to identify specimens

found 'in the field'. In an anthropological sense the table facilitates the task FIGU R E 5 . 1 The Identific
of identification in an interesting manner. Thus, one can see that the table is Source: Adapted from
ation of Fungi
M Jord an The EncycIopae .
dla of Fungi of Britain and Euro
based on little more than a series of binary divisions. So, either something Newton Abbot: Davi
d an d Cha rle� . pe, 1 995,

78
79
 D O I N G TH INGS WITH DOCUMEfns
LINDSAY PRIOR TEXTS

At a more detailed level wedirecwould also be able to deteritmine that the text urges TABLE 5.1 Prevalence of psychiatric disorders in private households by

people to act in specific Mol,tions . In parti cular , enc ? urages people to gender. Rates per thousand of population in past week rGB 1995)

'perform' (GoHman, 1959; l, taste 2000) classi ficati on in a defm ed m��ner. Thus Nature of disorder

I m orde� to
Females Males

it suggests that readers smel , look, �pick nd ? rese

logic: al fung
ve
and progressIve
M ixed anXiety & depressive disorder (MADD) 99 54
identify and distinguish7)onesuc from the other usmg a 34 28
5.1. So as
Generalized anxiety disorder (GAD)

� check� . as a re out1i n � d in Fig � re 25 17

sequence of checks' (p. r With the Ident ity of an amat eur , the system
Depressive episode
14 7
well as providing the reade
--
All phobias

orders 15 9
of classification also provordeides a script for doing . In that sense , the text Obsessive-compulsive disorder (OGD)
9 8
its readers as much as it alpha rs 'things' (fungi) in the worl d. ' Panic disorder
4 4
es in th� gre.at Pans.
Functional psychoses*

In the same way, the reade betic order ing ? f entri 21 75

es -It orders
Alcohol dependence*

as much as It orders the entnsize, 15 29

Encyclopaedia also ordersbyitsletter rsorder r than by entry orulddate,
Drug dependence*

them to search the text (It is, of course, rathe an order ing process that w? be�n
Rates per thousand of population i n past twelve months.

(1 995)
or author, or topic link. le to be one that is both effici
*

and. convement.)
Source : Meltzer et al.

considered by most peop s that is happening with ourent�ungi encyclopaedIa

any event, one of the thingback on its readers and structurmg them as an::�teur
(in use) is that it is acting h and Woolgar (1990) for similar examples) . Ihere
field scientists (see Lync of ordering fungi, and other classificatory syste
are, of course, other waysbe linked into different t �hn?logic. al systems ms.
They, however, would l groupings. Thus, a claSSI�fIcatIOn of fungI. ac�ordm �ndg
thereby into different sociagenetic (DNA) code would reqUl.re a very �Ifferent
to, say, the features of the different skills, different ways of workmg and
group of readers -- with ological hardware.
alternative arrays of technthings, how they make thmg . VISIbl . e, a�d how such
How documents place into sOcial practic s can s£�rm
.
is intended to be used, together with others related to it, as a resource for
researchers. Thus, we can, for example, refer to Table 5.1 as evidence for our
statement that, in Great Britain, about 16 per cent of people in any one week
show symptoms of a neurotic disorder. But how were this and other facts
arrived at? How was the report put together? And what would an answer to
such questions tell us about fundamental social processes?
The study of the manner in which social data such as are found in Table
�.1, �re pr� duced, . has a good sociological pedigree. Some early markers
In thIS partIcular hIstory are available in the works of Cicourel (1964, 1976),

systems of visibility are tiedd, it is parti. cularly �mstruchve toa gUldm
for social research. Indee s of representation can n::aterializeexam
graphical and similar formwise remain opaque and dIffuse. (For thing
nomena) that would otherrally, in order to make things visible, huma
see Prior et al., 2002.) Natulate ideas i to images and traces. S�ch proce
and agents have to trans and what�IS supposedly the ,same objec
of translation are various,into a number of alternative forms. How the formbes
translated (Serres, 1995) how they act back on their creators is, however,
relate one to the othertoandempirical research.
always a matter open issues of representatIon . m. a manner that IS,. perhaps,
So far I have discussed us social scientific c nc�rn
slightly detached from obvioon a form of classi. ficati ? on/s.repre
I intend to focus specifically
commonly used to provide a description of social behaviour.
T H E NAT U R E OF N E UROSES
. g theme Douglas (1967) and Garfinkel (1967) who tended to focus on crime and suicide
. how
me stat�st�cs. As with the manufacture of crime or suicide or any other form of
s (phe­ statistIcs, the production of psychiatric statistics depends on the existence
an example, of at least two foundation stones. The first underpins what we might call
n actors a concept� al �r theoretical scheme, whilst the second underpins the rules
and techmcal mstructions for applying the concepts to a set of events and
. t cansses
occurrences. Now, the conceptual scheme in terms of which mental illness is
compreh�nded is that contained in the Diagnostic and Statistical Manual of
the Amerzcan Psychiatric Association (the DSM IV, 2000). The latter contains
a �eries of c�tegories relating to the various psychiatric maladies that people
�I�ht bethesaI� t? suff�r from, and it also contains diagnostic criteria for recog­
So in the n�xt sectio� mzmg dIstmct dIsorders. Some of the disorders are listed in Table 5.1.
sentatIOn that IS We will return to the DSM in a moment. For now let us focus on how the
figures in Table 5.1 were obtained.
The data in the table are derived from answers to a survey conducted
ove.r 10,000 p�ivate households. The appendix to the research report fromin
whIch these fIgures are derived provides the detail of the sample frame
that was used (Meltzer et al., 1995). It tells us the rules by means of
hous�hold� and �he adults within them were selected. It also provideswhich

I am looking at a table of research results (Table 5.1). It tells us about the
community prevalence rates in Great Britain of ce�tain types o� what are
sometimes called 'neurotic' disorders, together With some estlmates for
the 'functional psychoses'. The table represents 'facts' about mental illness. It
80
que�tlOnnalre or mstrument by which mental illness was recognized. In this the
particular case the instrument was called the Clinical Interview Schedule
(ReVised) or CIS-R. The CIS-R is one of a variety of 'instruments' that produce
81
 �
D O I N G THINGS WITH DOCUME TS
LIND SAY PRIO R TEXTS

clinical and other phenomena (see,-Rforareexa
respects documents such as the ClS �uch as
ducing 'things' . Indeed, phenomen�
illness and 'quality of life' are condItIons
ments of the kind referred to here. In the
mple, Bowling, 199 7). In many
lik� �ac��ne, tools tools for. pr�­
�IsabIlIty , types of psyc�Iatnc
routme1y manufactured by mstru­
case of the CIS-R, the tool oper.ates
con�itions is made in the interview schedule and certainly at no stage were
subJects asked whether they suffered from obsessive-compulsive neuroses
or depression or whatever.
Human be.ings, of course, rarely pick up disorders singly and sequentially
and . an easIly labelled fashion. More likely they suffer from many things at
111

through a system of questions and ansing wers. For example, th.ere are q�estI�ns once - they are both depressed and psychotic, say. They indulge in substance
abuse and suffer from anxiety. So given that many individuals suffer from
about people feeling fatigued or feelyou've ill. One such question asks, Dunng multiple pathologies at one and the same time, the CIS-R provides precedence
n lacking in energy?' Another
the past month, have you felt that sort of bee e or pain in the past month?' rules that enable mul.tiple �isorders to be placed in a hierarchical sequence.
question asks, 'Have you had any 'Yes' orach'No' . For example, depressIve epIsodes always rank above phobias. In this respect,
Respondents are required to answer Sociology, Cicourel (1964) dIsc . ussed the
In his Method and Measurement in latter raise� issues relating to the ,:ays
the report follows the system prescribed by the DSM.
It was by the posing and coding of questions and answers, then, that
status of questions such as these. The the questIons and answers provIded diagnoses of psychiatric disorder were arrived at. In fact"there is a sense in
in which the interview process, and edded. In particular, he became inter­ which one could ar�ue that the data assembling process - using interview
within the interview, are socially emb interview - as a social eve�t - can turn schedules and algonthms - manufactured the disorders. So what does this
ested in the process through which the, pointing out that the use of mstruments �uggest about our concept of psychiatric disorder? And what would happen
conversation into social scientific data act of measurement by fiat. That is to If the research managers had set the cut-off point on the schedule to 10 or
(such as, say, the CIS) involved anmonality of meaning on q�estl. Ons and 18 rather than 12?
say, the instrument imposes a com iously understood - at dIfferent tIm . es �uch questions are in many ways related. It is possible, for example, to select
od, var
tkoup-St�enstra, 2?00). Indeed, the ,whole
answers that are, in all like liho a dIfferent cut-off point. Moving the point to, say, 1 0 would increase the
and by different people (see also Houscie nces IS somethmg of a Pandora .s b.ox, prevalen�e of mental illness in the community. Moving the point to 18 would
matter of measurement in the social whe t�er or not the C�S-R (or a sI�I1ar decrease It. SO we can have �s much or as little mental illness in the community
raising as it does problems concerning thmgs as depressIOn �nd a�x I�ty. as we. want. (It was once saId of Poland that it was a country on wheels seeing
instrument) is valid for measuring sucthissu es at the level of the mtervlewm? how Its borders were changed so frequently, and one might be inclined to
Rather than concentrate on measuremen useful at this point to tur.n to an e�am �­ take the same view about the prevalence of psychiatric disorders.) There are,
process, however, it would be more ced s by mea ns of whI ch psy chIa tnc of course, conventions about where the point should be, but the fact that
nation of the computer-assisted pro to ure focu s on the rules-based procedures the point is movable tells us something about what Foucault (1972) would
diagnoses were made. Tha t is to say,
by means of which the research eman agers mo.ved f,rom th� 'Yes� /'N�' have referred to as a particular type of 'discourse' on mental illness in the late
responses on the interview schedul tosIve t�e ca�egone� of defress�ve epIsode , twentieth century. It is a discourse that argues that mental illness is not
something that is qu�litatively �ifferent from sanity, but rather something of
'panic attack' and 'obsessive-compul dISorder contamed m the tables the same orde� that dIffers only 111 degree. The implication is, then, that states
of the report. ' interviewers. I� other words,. the of health and Illness can be arranged along a continuum - a continuum that
The CIS-R depends on the use of 'laymoo d and behav.IOur are not tramed runs from zero to infinity. Whether or not individuals are to be deemed 'ill'
people who ask subjects questions aboossI ut. depends not on what they think, feel or do, but on the cut-off point that we
psychiatrists. Consequently, it is imp comble for them to diagnose the respon­ �se for our classification. The level at which the cut-off is set is important, even
ed interview schedules t.o the
den ts. Instead, on the return of the e putplet into play. Ma�y of the �l?ont�ms 111 research terms, mainly because the impact of such things as genetic or social
research team a set of algorithms wer ple, the stionnaIre wa.s dIVIde? mto factors on psychiatric disorder can be amplified or even ' eradicated' by moving
involved scoring responses. For exam each sectque ion. On the sectIon relatmg to the point upward or downward (see, for example, Brown, 1981).
sections and points were awarded in score 1 if they had been 'generally This vision of psychiatric disorder as a quantitative variation on normal
anxiety, for example, a respondent wouorldmor e days in the past seven days. behaviour expresses only .o�e of a number of possible positions on the subject.
anxious or nervous or tense' for fourtense, nervous or anxious for more than It was, for example, a VISIOn that used to be contained within the DSM.
They would score another 1 if they feltpast seven days - and so on. As we move However, the contents and the theoretical ideas behind the DSM have changed
three hours in total in any one of the th� points, and if they total more than m�r�edly between the appearance of the first (1952) and the fourth (1994)
through the questionnaire we add IS assumed to dIsp . lay sympto�s of a :�Ibon. And o�� consequence of the alteration of the conceptual scheme
12 (the cut-off point) then the subjectit is, is dependent on whIch s�ctIOns �f IS that the condItIons that we are referring to in the 1990s are not the same
psychiatric disorder. Which disorder within. So, no mention of the dIagnostIC conditions as were referred to in the 1952 edition. For example, the word
the instrument the respondent scores
83
82
LIND SAY PRIO R G TEXTS

ure , course and origin of that depres sion

n' occ urs in bot h, but the nat
'de pre ssio turally, our table of statistics eleble (Ta
�
rad ica lly alte red (He aly , 199 7). Na erth ss :2
hav e
change, but the changes are nev
U

5.1 ) would not highlight this win

o

g trends in the prevalence of 'depre ssion' o

o
(f) CJ)

embodied within it. (A table sho th cen

-J U

tury would, however, be affe , in a « g, «

-J
cted o
0

over the later half of the twentierations )

fundamental way, by these altethe DSM, . ed
tory of then, we can see how it is produc
By examining the his entation - in a politically structut has been red spa ce
as with all forms of 'expert' docum fact, the DSM is a document tha ce of the
(see Rose, 198 8; Young, 1995). ]nexp
__

faction. Given the significan se, that

produced by a professional oration ert the global network of experti
American psychiatric Associ whatinis and what is not a psychiatric disorder
faction has the power to decidedefined. More importantly, perhaps, we begin
and how that disorder is to be chine tool - a tool that, when assembled with
to see how the DSM is like a ma te new products. Such generative documents
others of the same ilk, can genera which experts think and talk and write. In
set out the boundaries in terms of chiatric disorder, they are not algorit , of course,
the manufacture of data about psy CIS-R, the coding frame, the ether suc hms
the only machine tools at hand. The serve in the workshop. Put tog for ever a
h
to which we have referred, alsothe production process unfolds is
instruments produce, and how ch. For it is on the basis of such representations
matter of legitimate social resear ts' about society.
that social scientists refer to 'facchers tend to use the results of such research
Naturally, most social resear ource' . However, what the above analysis
as is embodied in Table 5.1 as rth 'res 'topic' - on
suggests is that it is alw ays wo while investigating such data as is to say, it is
an and Pollner (19 71) . Tha , pro t
this last distinction see Zimmerm ents are produced; who, exa (Forduc ctly ed
forever beneficial to ask how docum zed
tion process socially organi course, it is ? oth er
them; and how was the produc g produced documentation, of
examples see Prior, 2003.) Havin l use it. It is to issues of use that I shall
inevitable that someone somewhere wil C\l
-c -c
C\l

turn in the next section. \I: \I:

s:::. u s:::.
Q. -- Q.

:.c
o o 0
N C N

1.1)
C\l .-
b s:::.
o C\l 0
TI O N
D O C U M E NT S I N AC 0.. 1.1)

only produced, but also, in turn,92) are prodUC­

Text and documentation are not of a psychiatric record, Ha k (19 provides
tive. For example, in his discussion al psychiatrist translates item s of patient '"

an illustration of how a profession 'iii

r::

a written record. In so doi ng the psy chi atrist o

talk and observed behaviour intoess 'E

'"

ential details of a patien t's con ver sation,

as note taker highlights the ge
_

delusions, hallucina tion s, dia gno stic �

codes them into professional languas for(offuture action (entry into a psychiatric
_

terms, etc.) and makes suggestionwe can see similar processes at work. Figure Lfi
C\I

unit or whatever). In Figure 5.2,nursing assessment notes that I came across LlJ

5.2 is a facsimile of a page of late 198 0s. It is clear from the notes that the a:
:::l
in a psychiatric hospital in the cerned to categorize their patients in a
CJ
u::
members of the nursing staff were con
84
 D O I N G THINGS WITH DOCUMEthS

L IN DSAY PRI OR TEXTS

. nc I u ded a one -wo rd dia gno

sis of the patient'ses text. Thus I provided in my study (Prior, 1993) a short extract of an exchange
wa ys. Th e lat ter
variety of assessment of ;act1. l. fles f daily living' (ADL) skills, brief not between nurses in the ward office concerning the issue as to what was wrong
condition, an level of cooperatl:'On an� hygiene exhibited by the patient and with 'X'. Thus:
concerning the ese ass ess m ent s we re b as � d on con ver sat ion s and interchanrecges
Th
so forth. s and 'patien
rse . ts n nur ses alo ne. It is important toulgall N urse 1 : Does anyone know what's supposed to be wrong with X?

between nu
ien ts/ ch n a � � � � l � um an bei ngs com monly ind ed [Blank looks and silence meet the question]
that the ay of act�1Vl:�tl�S, a�;��haviours. For example, patients would talk
pat N urse 2 : Schizophrenia, I suppose.
_

in a wide arr watch te eVl. sIO. n, 1e;d each other cigarettes, shout, laugh, Nurse 1 : Hmm. I 've never seen any sign of it.
to themsel:es,the hospI. tal work��o and so forth. Yet of such a myriad array N urse 3 : Well , he's on chlorpromazine, so he must be schizophren ic.
go to work m y a few are ever 19�.i?' hted . Thus' in the case of the first named
of activities
. onl , rem . a' the1r poor ADL skills, their temper tantrums The MCD of 'schizophrenic' recorded in the notes is, then, sustained and
, th '
elf SC l1 '
1 Z oph
patient, ome behavl. Our tha' t are highlighte d. This selectivity of focus
It IS underlined in this case by means of a casual conversation - and especially the
and quarrels e even more eVI. dent were one t examine other kinds of patient 0
reference to 'X's' medication. (The use of medication t9 define a specific
would becom in the hospl. ta1 : Wh' h I am currently referring there were
0
psychiatric disorder, rather than the other way around, is not uncommon in
records. Thus,ic records (calle� char;:) and social work records kept on each psychiatry.) If, however, there were any real doubt about 'what was wrong
also psychiatrformer ere rr; l�tam. e� b th medically trained psychiatrists with X', it would be the notes that would carry the day. So the researcher
patient. The other km � ds �m ormaf o�, su�h as data on whether the patients
0
who wishes to concentrate on the use of documents in action has to be
and.con. tain ed k sym pto ms 0 � ( f s ChI ' zop hre nia)' their me. Ons dication constantly aware as to how the written record is tied into and anchored within
e d any 'f lr
exhlblt s, the�lf ,hls. tory, lte t ran . ms about family life, patient deluSl , and other aspects of organizational life such as conversations at the nursing station.
and its eff ect hla. tnc'. records I ked very much like those alluded
00
Nevertheless, it is only when assessments are written down and can be pointed
so for th. In fac t th e p syc Soc ial wo rk rec ord s were also :nad � up to that they are used to form a foundation on which routine social actions
to and rep ort ed u on by Ha k
�ients. These aid relatively little attention to medl� . andg­
(19 92) . a l dla are built. Thus medical professionals can and do use 'the files' as a warrant
for eac h of the pat . atlO � n and referred more often to the stabIlity for their actions in relation to their patients - showing how what they do to
noses an d t h ee ff ec t s o f me dlc . relatlO' nsh'lps wl'th others, the nature and leve1, patients is warranted by the information on the record(s). Indeed, in the context
ma tun ty
·
0f t h e pa f en t viS -a-V IS
: s, and the lI" ke CO�Sl. derable reference to the whereabouts of psychiatry Barrett (1996: 107) has underlined how 'clinical writing' not only
of the ir sta te ben efit er 'famIly members was also made wit. hin social describes the treatment of patients, but also constitutes the treatment.
beh avl Our s an d opm ' l'on s ofo th make entnes m' profes- such In a wider context, Bowker and Star (1999) and Young (1995) have also
such rec�rdS anthe d the 'right' toof the
·

work files. Accessre orto les s restncted. t, ;l.�t'memb

0 ers ual
individ part,
.
pointed out how (documented) nomenclatures of disease are routinely tied
records wereing mo pec t t� e SC in eac h doc um ent served, m into the financial accounting mechanisms of hospital life. So, psychiatrists
sio nal gro up s. In tha t res s parties� Social alworecrkord talk routinely draw down a diagnostic category from the DSM, 'fit' it to a given
ma rk out the rea lm and exp er Ise � . e
. var iou s, patient, and then justify what was done for and to the patient/ client in the
to
ong ed in soc ial wo rk rec ord s, !, syc la t nc t a lk bel onged i medic light of the category. (This, in the full knowledge that categories and patients
bel 1 ong�d in nursm.g record
s. .Ive. In
and nurse talkobsbeerv atI Ons on pat I nts / clie nts are , the n, hig hly sel ect rarely make a 1:1 fit.) Indeed, the naming of diseases and disorders in modern
Recorded
the case of publicrefserervic � agen�J !�l��d sU�hticul records often def
ar
ine the humanl
ways. In so doingdevtheicey scalor
medical systems is often used more for purposes of financial reimbursement
(to and from insurance companies), and other accounting and monitoring
bei ng s tha t the y to m spe i P a tion purposes, than they are for forming accurate descriptions of a given patient's
upon and actlva . te a whole sen. es f. membership categoriza
0
be ass ociat ed WIt. h condition.
n, 199 Ho w a par tl u a : devi ce com es to
� atiOn might be used and called upon
MCDs (Sil" verma 8).
I and how that categonz
1 Arguing along a similar path, Zerubavel (1979: 45) had previously indicated
any md1Vlduaand
·

lam . �n individual's behaviour in specific circalum stances how notes written up by medical and nursing professionals were ' among the
to account foroccasiexp ndamental socve, iologic .earch.
res main criteria used by their supervisors to evaluate their clinical competence',
can form thork e . On for lm portant and furef
al erred todiffe abo I was pnhrou man'1y as well as forming the primary mechanism through which continuous super­
In my w. on the f SYch'latric hosbepitcla ed in rent and recordghs
wa ys t vision of patients was maintained. In fact Zerubavel's study highlights
interested m how patients. came to3) N turssifi ally the use of notroues nds p tients. the centrality of charts, graphs and records of all kinds in underpinning the
routine procedure s (see rno�� 199 fic:tion system tha t sur routine social organization of hospital life. Thus, printed schedules are used
forms only onever supp�rt or casideualntlint
sat l�n a n � erc han ges for m ano the r, and�m anyd to organize the patient/ staff day; printouts of various kinds are routinely
Everyday conthere wlll always be a constant interchange between talk an used to monitor patients; and notes written so as indicate how the 'hospital'
organization 87
86

LIND SAY PRiO R " TEXTS
rse, patient records are, as we have
cares for its clients. (In US hospitals, ofkcou ch to hang financial costs �nd
already indicated, also used as a hoo seeoninwhi the use of docum�nts to rr:omt�r
transactions.) Above all perhaps, we self-reflexive capacIty that IS saId
activity an essential component of that ern�ty as a .whole. .
by Bauman (1 991 ) to characterize mod . ds
. .
' .
Iate SOCI al rela tlOn shlp s of all kI �
This capacity of medical reco rds to med
Ber g (1 996 , 1997 ) who poin ts out how hos plta
has been further researched by records an.d acce � sse thro ugh rec ord s.
patients are both structured through tenc IS emp haS Ized by Berg IS. the
One important feature of pat ient exis e that
d so as to keep the case (and the
manner in which medical records areofuse ent tr�jec�ories �hrough records
patient) 'on track' . Such structuringningpati momtonng bemg two of them.
is achieved in numerous ways, plan e theand column headed 'Plans' in Figure
In this respect it is of interest to notmost imp ortant plans concerned whether
5.2. In the context of these records the in the community. (My )W� hospital
or not the patient was ready for life ortant cusp where psychIa �
tnc patIents
study was executed at that rath er imp
into 'the com � m nitY ' · m?st case�,
) In
were being moved out of hospitals and (kept in the hos pIta l war d). lhe . detml
the patients were to be 'ma intained' orta
hardly concerns us here. What is impon. Oth nt to note is that r � s of thIStm�ype
cord
rules for actl On are co �1 ed
always contain some rules for acti Latourer(1987) has used the term �actlon­
in the column relating to medication. ns written down in one context and
at-a-distance' to indicate how decacti isio
setting can carry implica tion s for on in future settings. And it i� in�eed
the case that records often contain inst ructions for future orgamzatJonal
chapter provides an excellent
activity. King James's claim that opened this
reflection on the process. the only things that are kept 'on track'
Patient trajectories are not, of course, are als? constr� cted through docu­
by the use of records. Patient identiIt�i�ISs of conSIderable mterest to study how
mentation and in the modern world red by various forms of identification.
our identi�ies are supported and alte provided some detail�d exan:ples as to
In a larger study (Prior, 2003) I have rrel In the followmg sectIon I draw
how self and documentation are integhts.ated.
upon just some of the attendant insi
TEX T AN D
Morgan examined the social nature
In a paper published in 1993, Miller and indicated how such c.v. s can be
of the academic c.v. In that paper theyhical practice - a form of practice that.
understood as a form of autobiograpenta of self in occupational settings
is centrally concerned with the pres seektion
to demonstrate how the acad emic
Using Goffman's (1959) work they of fron
and manufacture of self. Above
c.v. involves matters of presentationacademit c c.v. is a matter of performance.
all, however, they point out how the rred to are, of course, executed solely
Performances of the kind just refe of special concern to us. It is also of
through writing. In that respect they are
88
D O I N G THINGS WITH DOCU MENTS
interest .to note that Miller and Morgan indicate how the construction of a
c.v. can Itself engage the use of generative documents of the kind mentioned
in the �hir� section. In this instance the generative document referred to was
the Guzdelmes for the Presentation ofa Curriculum Vitae of Manchester University.
Ra�her than look at how human actors can use and manufacture documen­
tatlOn for the presentation of the selt however, I propose in this section to look
l at . how d ?cuments can structure their readers. That is, to look at how they
mIght gUIde readers to perform in specific ways.
Ir this context it is helpful to note Iser's (1989) claims about how the reader's
role :m the.work of English fiction altered between the eighteenth cenhl�y and
the twentleth. v:'e nee� not attend to the detait but simply note examples
of tl�e ma�ner m whIch texts provide the reader (and the writer) with
speCIfIc. scnpts for performance. For example, Iser point1} out how fictional
texts have structured the reader as passive recipient of a narrative or - as with
�hackera�'s Vanity Fair as an observer of life's variety and display. With
Cha:les DIckens, on the other hand, the reader is almost in variably structured
as � Judge o� moral character -- this, now here more so, perhaps, than with Oliver
TWist. In tl?S :ext we are i�veigled into taking up a moral stance on poverty
and explOItatIon, subserVIence and domination, whilst the writer remains
positioned as a dispassionate teller of a tale. That stance of reader and of writer
is, of course, quite different from, say, the one adopted in Dashiel Hammett's
The Maltese Falcon, a story �hat features the unforgettably amoral Sam Spade.
. works of fiction but with routine forms
Naturally, our concern IS not WIth
of docu��ntation. In that respect, we have already glimpsed at how a work
of non-.fIctlOn (the encyclopaedia of fungi) can structure its readers. There we
noted how the reader was structured as an amateur and how that amateur
was directed to identify fungi in particular ways. But even with everyday
documents re� ders and writers are structured and allocated to specific roles.
To that .end I mtend to refer to a form of documentation that is provided to
people m the UK when th� y collect medicines from a pharmacy. In each case
where that happens the clIent should be provided with a Patient Information
�eaflet (appropriately abbreviated to PIL). For each form of medication there
IS also a 'data sheet' for pharmacists. It will serve our purposes to analyse the
text for both forms of document. We will do so with respect to leaflets that
relate to one and the same drug.
The PIL to which I am going to refer is that for ziduvodine. The latter is
commonly prescribed (under a trademarked name) for people who show
s�m�torr:s of HIV in.fe.c�ion and AIDS. The PIL for ziduvodine opens by
hIghl.lghtmg the possIbIhty of the taker of the medication having doubts and
wo:nes . Should the 'patient' have any doubts it is suggested that they consult
1 heu doctor . And this pattern of the doctor as an expert (with full know­
.
edge, but WIthout doubt and worry), and the consumer as in need of advice
.IS replicated throughout the leaflet. Consequently, the doctor is presented
as s?meone who will assuage doubt and worry, someone who can provide
adVIce on all aspects of medication, and as someone who will give instructions
that ought to be followed.
89
 D O I N G THINGS WITH DOCU MENTS
L I NDSAY PRI OR % TEXTS

t
, worry. . . . If you take a larger dose than prescribed instruct us as to how we should proceed with our research. Most of these
know as soon as possible if this
e a dos e, don _

to tak happens . . . . If texts focus on ways to capture and analyse speech and thought and behaviour.
. . . you shoul d 1et
If you forg et
you r doc tor
. b mistake tell your doctor at once . . . , Y ou However, few social scientific research manuals concentrate on the written
�
.lcme
less yOU do�tor t�lls you to.
someone else takes your med (ABPI, 1 996) word and, more specifically, on documents that contain words . Indeed, when
should not stop treatment un documents are put forward for consideration they are usually approached
. urm . g of the p atient as a non-expert, the leaflet also in terms of their content rather than their status as 'things'. That is, the focus
In line wIt h the str uct . w1lIC. h the drug acts on the body.
nar rat . 0f the manner m
lve is usually on the language contained in the document as a medium of thought
Provide . . s a lay . a1 a gent that is ziduvodine serves to and action. Yet, as I stated in the introductory section, it is quite clear that
ve m d'l at es how the an t '
I-vu
The narratl
pro gre ssIO
�
n of EIV an d AID S but does not provide a 'cure'. Instead each and every document stands in a dual relation to fields of action . Namely,
delay the ord'n t the leaflet' EIV - leftint unchecked as a receptacle (of instructions, commands, wishes, reports, etc.), as an agent
it 'fights' against EIV . For, acc calle� C�4 cells and tur n the m o a 'mini­ that is open to manipulation by others, and as an ally or resource to be
- will enter into a group �f cellsg�� t st u gIe against EIV the ant i-vi ral agent mobilized for further action.
factory' for infection. But m thece sa an� s�op the factory �roducing the viral As a receptacle of content, one set of questions that may quite justifiably
also seeks" to invade the CD4 oic narrative . .) be asked by the social scientific researcher concern the processes and
agent. (It lS, 1f you 1'1
llze, an her hat . erent
dIff
. p h arm aCl' S t s, how eve r a som ew circumstances in terms of which document 'X' has been produced or manu­
In the respective data sheet for read 0f the fact tha't ziduvodine is 'phos-,
.
. can . factured. That is, to treat the document as topic rather than as resource. Yet,
narrative IS' provi' ded . There. oneate d env ' a t ' e by cellular thymydme kmase ,
IV documents are not just manufactured, they are consumed, and as with all tools
to th e mo no P hos ph
Phorylated as an mhI'bl'tor of, and substra te for, the they are manipulated in organized settings for many different ends . They
that it is Icatalyse d'. , �nd th�t it acts
I

and expect
•

. Nat urall one would both ly from that for h ope also function in different ways - irrespective of human manipulations. In short,
viral reverse transcnptasecists to rea�' s�1newhat differ ent documents have effects. So a further route of analysis for the researcher is
the data sheet for pharma . plex I11formation, for example. Butchem-
what to ask questions about how documents function in specific circumstances .
patients - so as to provide moretcom t in using this comple x lex icon of Naturally, the way in which a document functions is often affected by its
is important from our standpoinf is. tha , texts are also structunng
.
content, but content is not always determinant . Indeed, we need to keep in
istry, rather tha n meta phors � ��� �t:�cturi�g of expertise is also evident
l f tori es the
mind that the content of a document is never fixed and static, not least because
experts and non-experts. An .
. ph,armaClSts' dat . a sheet talks, for exa.mpele, documents have always to be read, and reading implies that the content of a
in more duect ways. Thu.s thelen . d h ow patients ought to be cautIOndder, document will be situated rather than fixed. In any event, the analysis of
about the management of pat th tsAll, an
I

' � a11 then the PIL structures the), rea content, production and use form three of the corner points around which
advised, monitored, �nd so for d'uec;IOn �f others (who are experts whilst we can consider the development of a research strategy. In this chapter I have
as an individual subject to the patients. sought to provide some examples as to the kinds of issues that are implicated
et structures the reader as an exP ert in control ofbeen
the data she . purpose m . 100king at the exampIes above has to indicate . in such a strategy.
My mam olves doc entatIOn.
ntit y wo r� , as per f r anc e inv um At the risk of seeming to be as egocentric as King James, with whom
ay � � the form of full-length books
of ide
how a large part �
In some cases that doc�menta � on r ordi:ary everyday life it usually takes
we opened the chapter, I am going to end this excursus by quoting myself.
My aim, of course, is not to hear the sound of my own words, but to argue for
- biographies and autobIO gra�s:�bl�n and w :iting a C. V., or filling out a job a redirection of social scientific interest in documentation as a whole. My
less not able forms - such as as ?
for a driver's licence or a passport. argument runs thus:
. . , a census form, an appl'lcatIOn
applIcatlon
Yet, exact1 y h ow P. eople use and ma mpu1 ate suc. h routine forms ofm
.
.
doc
. If,
u-
. a 1 egl' tl'mate field for resear earcher,
ch ltse It might be objected of course that the most important feature of documentation
mentation - to d0 1 de�ft1 y woryk0-f IS
-­

1oraf1 on by the qualitative res

content - has been given too Iow a profile throughout the entire [work], I might
and certainly one that IS worth exp defend such a lack of focus by saying that there are numerous other texts that deal
with aspects of content analysis, or discourse analysis, and that there would be little
point in duplicating such work here, However, I offer no such defence. There is no
need, Instead I offer an analogy. My analogy concerns an operatic libretto (the set
CO NC LU SIO NS
of words and phrases that is sung), Taken on its own a libretto rarely adds up to
, . ' �ts C mmonl lace on human actors manifests much, The text as narrative is often disjointed, repetitive and lacking in depth.
The emphasis that �oclal sClentl !�
itself most clearly m the attee.nt�nnd�sath ld we wish to study human actors
the ve to what such actors say and I cannot think of a single one that might hold a person's attention as a gripping
tale, Yet, a libretto is not intended to be analysed in isolation, It demands to be
think and believe and. opi.n, manner��er e are many manuals available to analysed in action, How it is integrated into the dramatic action on stage, how it
in a rigorous social sClentlflc
91
90
 D O I N G THINGS WITH DOCUMENTS
L I N D SAY PRIOR " TEXTS
Berg, M. (] 997) Rationalizing Medical Work. Decision-support techniques and medical
music, how it is called upon (recr� ited)
relates to the melody and rhythm of the practices. Cambridge, MA: MIT Press.
it is performed - all of . these are of pnmary
d mani ulated by the singers, how Bijker, W.E., Pinch, T. and Hughes, T.P. (eds) (1987) The Social Construction of
� �
portan e. Its substance as displayed on
the inert page I S of only secondary Technological Systems. Cambridge, MA: MIT Press.
concern. (Prior, 2003: 172-3) Bowker, G.c. and Star, S.L. (1999) Sorting Things Out. Classification and Its Consequences.
Cambridge, MA: MIT Press.
Bowling, A (1997) Measuring Health. A Review of Qual ity of Life Measurement Scales,
Second Edition. Buckingham: Open University Press.

pen. Scotland under James Vl and l. Urbana:

Brown, G. (1981) 'Etiological studies and the definition of a case', in J.K. Wing,
P. Bebbington and L.N. Robins (eds), What is a case? The problem of definition in
1 . Cited in Lee, M. (1980 ) Government by
psychiatric community surveys. London: Grant McIntyre. pp. 62-9.
Cicourel, A V. (1964) Method and Measurement in Sociology. New York: The Free Press.
University of Illinois Press. p. vii.
Cicourel, AV. (1976) The Social Organisation of Juvenile Justice. London: Heinemann.
Darnton, K (1984) The Great Cat Massacre and Other Episodes in F.ench Cultural History.
New York: Basic Books.
Douglas, J.D. (1967) The Social Meanings of Suicide. Princeton, NJ: Princeton University
Press.
Douglas, M. (] 966) Purity and Danger: An Analysis of the Concepts of Pollution and Taboo.
London: Routledge & Kegan Paul.
Foucault, M. (1 970) The Order of Things. London: Tavistock.
Foucault, M. (1972) The Archaeology of Knowledge, trans. A Sheridan. New York:
Pantheon.
Garfinkel, H. (1967) Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Goffman, E. (1959) The presentation of the self in everyday life. New York: Doubleday
Anchor Books.
Gould, S.]. (1989) Wonderful Life. The Burgess Shale and the Nature of History. New York:
W.W. Norton.
Hak, T. (1992) 'Psychiatric records as transformations of other texts', in G. Watson
and KM. Seiler (eds), Text in Context. Contributions to Ethnomethodology. London:
Sage. pp. 138-55.
Healy, D. (1 997) The Antidepressant Era. Cambridge, MA: Harvard University Press.
Houtkoup-Steenstra, H . (2000) Interaction and the standardized survey interview.
Cambridge: Cambridge University Press.
Iser, W. (1989) Prospecting. From reader response to literary anthropology. Baltimore, MD:
Johns Hopkins University Press.
Jordan, M. (1995) The Encyclopaedia ofFungi of Britain and Europe. Newton Abbot: David
and Charles.
Keating, P. and Cambrosio, A (2000) "'Real compared to what?" Diagnosing leukemias
and lymphomas', in M. Lock, A Young and A Cambrosio (eds), Living and working
with the new medical technologies. Intersections of inquiry. Cambridge: Cambridge

n leaflets, 1996-97: London: Da aP r � �� . University Press. pp. 103-34.

ABPI (1996) Compendium of patient informatio st1 al Man � a f en aI Latour, B. (1987) Science in Action. How to follow engineers and scientists through society.
Diagnostic and Stat7 �
American Psychiatric Association (2000)
0

c Asso .
clatlO n . Buckingham: Open University Press.
American Psyc hwtn
Disorders. DSM-IV-TR. Washington, DC: . Levi-Strauss, c. (1969) Totemism, trans. K Needham. Harmondsworth: Penguin.
the social defm ltlOn of schl.zoph�ema. An
B ar tt R
::
(199 6) The psyc hiatr ic team and
tJ;rop� logical study of person and illness.
Cambridge: Cambridge um erSl P ss. � ? :� Lynch, M. and Woolgar, S. (eds) (1990) Representation in Scientific Practice. Cambridge,
MA: MIT Press.
M. Ward and K Howard. Lon on: ona an
Barthes, K (1985) The Fashion System, trans. Miller, N. and Morgan, D. (1993) 'Called to account. The CV as an autobiographical
ty. London. RoutIe d ge.
Cap e. . practice', Sociology, 27 (1): 133-43.
Baum , Z' ( 991) Intimations of Postmoderm
'

. t Mol, A. (2000) 'Pathology and the clinic: an ethnographic presentation of two

�0 996) 'Prac tices of read ing and writ ing: the constitutive role of the pahen atheroscleroses', in M. Lock, A Young and A Cambrosio (eds), Living and working
: 524.
Ber
�cord in medical work', Sociology of Health and lIlness, 1 8 (4): 499-
'

93
92
LINDSAY PRIOR TEXTS

ge
sections of inquiry. Cambridge: Cambrid
with the new medical technologies. Inter
University �ress. pp. 8?-102 . ' ds, K.
H In . (1995 ) The prevalence of psychiatric
Meltzer, H., Gill, B., Pethcrew, M . and ric as a
te households. OPCS Surveys 0f Psychiat
.

morbidity amon g adults living in priva

Morbidity in Great Britain. Report 1 . Lon
don : HM SO.
Prior, L. (1993) The Social Organization of
Mental Illness. London and Newbury
Park,

CA: Sage. Annette N. Markham

. L. (2003) Us ' Documents in Social Research . Lond on: Sage.
Pr�' or,
�
:n ra J Pill, R. and Hughes, D. (2002 ) 'Making risk visible: the
�� � �;:� � �
s � the �s:ssment of (cancer) genetic risk'
pn
i ;f , Health, Risk and SOCiety, 4

aid, and Other FIgments of the CIaSS1'fymg

(3) ' 242-5 8. .
'
Ritvo: H. (199 7) The Platypus and the Merm
Pres s.
Imagination. Cambridge, MA: MIT . . '
ageable indIvIduals , HIstory of the Human
"

Rose, N . (1988) 'Calculable minds and man Qualitative studies of the Internet are diverse. In the early 1990s, Julian Dibbell
Sciences, 1 (2): 1 79-200. ' published an ethnographically informed account and analysis of a rape in
Serres, M . (J 995) Conversations on Science,
.
Culture and TIme, with Bruno L atour, trans
.

0f M '
' h19an
. IC Press : cyberspace (1993). In a popular chatroom space of an online community, one
R La idus. Ann Arbor: The University . ' ge..
�
Silv�rm n, D. (1998) Harvey Sacks . Socia l Scien ce and Conv ersatwn AnalysIs. Cambnd member utilized a program that controlled the text-based actions of two
females present in the room. He then proceeded to write acts of violation
. . (1989) 'Institutional ecology trans1at'lOns
Polity Press.
Star S L and G· nese ' , mer, J . R
"
" and boundary involving these women. While the women were powerless to do anything
;bj�c�s: amateurs and profe�sionals
ogy,
in Berkeley's Museum 0f Vert brate ZooI e
except turn off their computers, their online characters continued to be violated
420. online in front of many other people. The event had serious repercussions
1907-39' ' Social Studies of SCIence, 19: 387- . '
Inventing Post -trau matlc Stress DIsorder.
Young, A . ( 1995 ) The Harm ony
.
of Illus
.
ions.
s.
for violated women and the community in general.
Princeton, NJ: Princeton Umverslty Pres
.

. ' . In the mid-1990s, Witmer and Katzman (1998) studied computer-mediated

Zerubave1, E . (1979) Patterns of Time in Hosp
University of Chicago Press.
. H . and Polln er , M. (1971
Zimmerman, D .
J.D. Douglas (ed.) , Understanding Everyday
pp. 80-103.
ital Life: A Socwloglcal PerspectIVe. Lon d on.'
, .
communication (CMC) users and the ways they compensated effectively for
d as a phenomenon , In the absence of non-verbal and para-linguistic elements of conversation. Using
Life. Lond on: Routled ge & Kegan Paul.
) 'The ever yday worl
. emoticons, users convey humor, irony, and other emotions, supplementing
the content of the message to enhance interpersonal connectedness.
In the late 1990s, Norwegian scholar Anne Ryen utilized the capacity of
the Internet to conduct a long-term and long-distance case study of an Asian
businessman in Tanzania. While she began by simply using the Internet as
a tool for extending her reach, Ryen ended up also examining how the Internet
influenced her personal and professional relationship with the participant
(2002).
In 2003, Camille Johnson archived and analyzed nearly 600 web pages
promoting anorexia as a lifestyle rather than a disease. She contends that this
pro-anorexia network relies on Internet technologies to build and reproduce
members' ideologies. Through cutting and pasting images and common texts,
such as the "Thin Commandments," these women are actively constructing
a global yet anonymous community, which appears to provide solidarity and
helps to justify their choice to be anorexic.
These are four brief examples of distinctive Internet studies conducted by
qualitative researchers. As a communication medium, a global network of
connection, and a scene of social construction, the Internet provides new
tools for conducting research, new venues for social research, and new means
for understanding the way social realities get constructed and reproduced
through discursive behaviors. This chapter seeks to illuminate some of the

94
 �
INTERNET COMMU N I CATION AS A TOOL FOR QUALITATIVE RESEA CH

ANN ETT E N . MAR KHA M

TEXTS

of stu dyi ng the Internet and lor using

;�g::�; research practice, particularly in terms of collecting information
pos sibilit ies as we ll as lim itat ion s r .
litative inquiry. .
. dIscursIVe. milieu
Internet technologies to augment qua 3 As a context of social construction the Internet is a umque
that facilitates the researcher's ;bil't t� wIt�ess � nd analyze the structure
.
of talk, the negotiation of meant: nd Idenht� , the development of
relationships and communitie lthe constr�ctlOn o� �ocial structures
as these occur discursively. W�� �:r the researcher partICIpates or simply
DEFiNI NG c

ated terms cyberspace and the We b

es the ass oci observes , the" lin gUIS. t'IC and socral
. structu es em�rgin through CMC
pr�vide the opportunity for researche�s to tr:ck and analyze how language
As an um bre lla term tha t inc lud the
refer to the actual network and �
(World Wide Web), the Internetputcan ers. Ma ny people use the Interneteminail,a
exchan ge of dat a bet we en com al bUIlds and sustains social reality.
seemingly straightforward wa y: sending and receiving personand ise and
n, dow nlo adi ng maps, vie win g me rch
ing pu blic inf orm atio a.. Whether conceptualized as a communication med mm, ' a global network
access
g pur cha ses onl ine , and gen era lly using the technologies for informere of connection, or a scene of social con �truc t"Ion, t11e Internet offers the qualita-
makin ernet can also refer to social spaces wh
tion gathering and transmission. Int oug h the exc han ge of tive researcher many means of ob.servmg an. d lor interacti ng WI' tl1 partICIpants
.
.

e thr m order to study the complex mt erre 1ahon of langu age, tecI1110l ogy, and
c
tur es emerg
-

shi ps, com mu nit ies, and cul is

.
rela tion in delayed time sequences. Theedreon
�
.

er in rea l tim e or culture. Regardless of the general f ;:�� � ed, one caI� �tilize the Internet
as a tool for research topics unrela�:;
t and ima ges , eith
tex
a lon g tradition of social interactio n and community development bas as a ternet speCIfIcally. (e.g., using
cap abi liti es of the Int ern et. In sho rt, the Internet can be perceivedgua ge the Internet as a convenient and anonymous means of gathermg · mformation
the work of social relations, a lan s the �:�:I�;�i�C
. �ocla� r
. enomenon (e . g.,
.

of tec hno log ica l too ls, a com ple x net on racial attitudes) andl or study the fnterne
set so forth. The way one defines and frame
system, a cultural milieu, andint et-b ase d tec hno logies, as studying the way a special interest group �1 sus �ms communIty .
with Int ern through the copying and pastin of grou speCIfIc Images m a network of
_ "

Int ern et infl uen ces how one era cts

well as how one studies the Inticern et. litative websites). Put simply, the Inter;et is botK a tool of research and a context
Which of these metap hor fra meworks is most useful for qua ­ worthy of research.
ear che rs? Wh at doe s the Int ern et con tribute to the endeavors of lita qua
dy, As with any metaphoric framework' th�se three frameworks guide and
res end on the specific phenomena und er stu ..
tive researchers? The answers dep and the methodol ogi cal app roa che s fav ored . naturally
. restrict the qualitative researcher s general appr oach and speCIfIc
. .
the res ear ch que stio ns ask ed, Internet is practIces m using and understandin? the Int ernet, allowmg the researcher
s can help illustrate how the che
The following three framework
typically conceptualized and the refore how the qualitative researnal stu r might
dy.
to focus on certain features �
views. Understanding the :n:��t��:��:. es 0 .these frame:vorks can help
; s at he expense of other p ossible
or stu dy it as a con tex t in itse lf or use it as a tool in a traditio researchers make wise choices as they. m�estIgate potentIally unfamiliar
use . stud Ies m which Interne t technologies
nnels for research environments or d eSIgn
n, the Internet provides new cha augment the collection or analysl's of m .
' formatIon.
1 As a medium for communicatio h other, new channels for researc hers to
people to communicate with eac new venues for conducting ear res ch.
communicate with participants,incand singly augmented with mo itional vin g and
Still primarily text-based but lsrea h parallel and depart from tradcursive
THE INTE R N ET AS A M ED I U M FOR C O M M U N I CATI O N
still images and sound, these tooearbot chers can tap into emerging discultural .
media for interaction. Thus, resdying I begin with the assumption that qualitative r�searchers ' analyze dIscursive
way people use CMC in
forms and practices, either stuinteract the h participants.
practices in naturalistic settings to . thenance
eIp bUI'ld nowledge related to. the con-
. I practIces
contexts or utilizing CMC to ernet wit col ses physical distances betwe en struction, negotiation, and mam of hum a n SOCIa and
2 As a network of computers, thepot Int lap
lab ora tion s structures . Whether explor'mg cu Iture wnt. large or a . s'mg1e conversation,
people, thus creating the net ential for collectives and colreach of the we can say that most qualitative inquir is' gro.unded m . mformation collected
not heretofore available. This le. The work extends the potential se world­ from observation text" talk and mtervlews
' ! (SIlverman " 2001) . A t a very
speed of transmission in the baSlC. level, then, 'qualitative res ear h ers engage m . the1997
researcher to a more global sca archiving capacity of computers, transforms
process of studying
wide networks, along with the ct. As individuals gain control over how communicative practices in contex/
time into a malleable constru s with others, researchers gain consider­ Inserting the Internet as a medium for int�ract�. on between researcher and
time structures their interaction conducting research. Understanding and participant or studying the Intern�t as It' m �dl�tes mteractions among subjects
able flexibility in designing andspace in creative ways can significantly in the field changes the researc scenano m that the Internet influences
utilizing time and notions of
97
96
U4

I NTERNET COMMUNICATION AS A TOOL FOR QUALITATIVE RESEARCH

ANN ETT E N . MAR KHA M

TEXTS

ultaneously mu ndane and be simply a new form of writing letters or leaving short notes for people
commumc. Iatl'ven ways , that are' sim
On practices inl not fea tur es of the without the scraps of paper, The Web might be a means of finding and
ne wil e SIm ' I an' tIes bet we en many
I
profound . E eadieasr0me ., . for comdmusonfl'orcatthion, cer, suc h as lett er wr itin g, tele- purchasing products and services without leaving one's home, Just as a
Interne - and ph PoSt-ItdIa
t
not es, an ties
tain capaci mter . and uses 0f hammer augments our physical strength, CMC can be perceived as a tool
that extends some of our senses out to a global level. For researchers, this
phone, te1egramum�catIon
, um' que!1 �ha.Pe a user's percepti ons and ' Ons.
actl
Internet com s extend al to the social and cultural; tool can be utilized with great benefit. Mann and Stewart (2000) provide
These influencethese combey o� : e mterperson ses have th� potentetialmetodIashitesft an excellent and comprehensive review of methodological and ethical
outcomes 0f prac Ices atmu mcatlOn , Ess pro ces considerations under the umbrella of Internet as To01,2 In addition to using
, cu1tura1 1eve1 , nsentand ially, the Internle outcomes
c
,

sensemakmg e ways mothe l'

ate, s -- m' teractiocultural lev the possib , Internet these new tools for exchanging information, interacting with participants,
- and m' somctlO, ns attheder els, Equally litative or collecting discourse, qualitative research usefully explores the tools
ofthese intera have the podytenadItla.clTOU Ph.'and
to s ft the wadatysa, in which qua
1
themselves as well as social interactions afforded by these tools,
2 As a place for communicating: Many users and researchers conceptualize the
technologies lect, make sense of, and represent Internet as a place as well as a tool. From this perspective, the Internet
researchers col describes not only the network that structures interaction but also the
cultural spaces in which meaningful human interactions occur (Jones, 1995),
From
't�I'nformation vir sly
tually instantanofeouthis, Internet interactions have no literal physical substance, yet they can be
t_
dm,m t�� tra�:X:� perceived as providing a visceral sense of presence (Soja, 1989) and having
The Internet is a me ters, mdlvld�aeco,mel a Jefin ple, Bec ause
rou s of peo stic of the Internet dimension, Novak (1991) tells us that once we discover space in informa­
between compunsm
,

information tracommuissnion has , fng characteri tion, we are freed from the constraints of architectures that occur in standard
ication IS 0 i'tted Hotedwewit
f ten con fl a h the channelstheorirmeVIe,diaw three dimensions. Internet communication can be seen as "liquid archi­
and the termI h messages nsm ' ver, by shifting atlO, n as a
th rough Wh'cdUlt, mo. del toare� tra rIghtl� different view of
communic issues tecture," which "bends, rotates, and mutates in interaction with the person
from this con cess 0f meanmsg-m mg, researc11erins fin d that other oven who inhabits it" (Novak, 1991), Interactions in these sensed dimensions are
contextual ,prot. The beauty 0f the .Intakern y which it is interw not merely meaningful but can have genuine consequences for participants,
become salIen making process at van' ?uset islevtheelswaInterfaces on the surface of as exemplified by the text-based rape mentioned at the beginning of this
chapter (Shields, 1996; D ibbell, 1993).
with the senseilitate certain interpretatlOns f th� medium' exerting influence In this frame, the Internet can be a tool but is also a location where one
the screen fac e user perceives the commumc� atI'on process . Below the surfac e,
on the way th the informatio, n excha ged lS' de sense of individually wlt, hm' can travel and exist and wherein one's discursive activities can contribute
directly to the shape and nature of the place, Researchers can take advantage
the content oftext, add·ing several van�ables t�;he complex relationship, among of these sensed dimensions to create interaction spaces that facilitate
a speCl'f'IC con technology, Ind'IVl. dua1 negotiation of this relationship mteracts particular types of engagement with participants, Alternately, and perhaps
self, other, and otiation pro��se� more present in the past decade of research in this area, researchers have
with others' neg elsewhe:e ;mr h/lVm�199 8' forthcoming) that people tenetd studied these sensed dimensions as cultural contexts (see, for example,
I have argued Int ernet m d'IS �ve �t asayasp, Som e conceptualize the Intern Baym et al., 2000; Jones, 1995; Kendall, 2002; Markham, 1998; Turkle, 1995).
to experience1the 1
, while oth" ers mperthcele worId These lace , Stil lizations resnceu1tthem'
l oth ers exp erie Using basic terms, one can study the space itself, the interactions within
as on1y a too wa conceptua techno1ogl. es,
Internet as a t Ys ofs ofbemandg mt ' ernet-based heuns. tlc, 0f
ct'lOns Wl'th Intcatego these places, and the relationships and communities formed through the
very dl'fferen a� m� d'lVl'duals' fallerainto these neat ries, this g h�W interactions, In my own work (Markham, 1998), the Internet is an umbrella
term for those social spaces constituted and mediated through computer­
Although not wa of bein is a useful g poi for con) con
nt siderin
tool, place, andextens�lOn, qual�ta lVe researcstaherrtins ten
1 l'
d to (or can ceptualIze mediated interactions, In addition, Jones (1995, 1999) provides several
volumes in which authors examine Internet as Place.
users and, by the Internet: 3 As a way of being in the world: One can also conceptualize the Internet as a
and approach way of being. In this sense, Internet-based technologies provide a means
.��tmg
" e ' ht nat ura lly or del ibe rate ly conceptualize for reinscribing, reconfiguring, or otherwise redefining identity, body, and
1 As a tool for com mu m
or
,
re � � �t
ne l and tra nsm itti ng inf orm ing
ation, extendeen
et a s a too self's connection with other. For example, a user might have two online
ch to connec� ;ith
the Int ern others enteringnsim and crossing betwFrom
one's physical reaal fiel ks ultaneously, personae with two distinct personalities and gender, Recently, scholars
multiple culturve serds,t or perfOrml' �g muCMlti�leintasa straightforward way as a have argued compellingly that the performance of self through CMC has
this per�pecti d�t,:lOn :o ������:��::�i; for communication, Email might
1
allowed transgendering to flourish, both as a concept and as a way of life,
convement ad 99
98
ANNETTE N . MARKHAM " TEXTS
INT ERN ET COM M U N I CAT
ION AS A TOO L FOR QUA
liTA TIVE RES EAR CH

because users can expenence . a different gender without the necessity of 1 Geographic dispersion
.
cross-dressmg, makeup hormones, or surgery (Future of Feminist Internet Temporal ma
2002). ThIso IS. a' good exam Multiple modallleaitybility
2
Studies,
. t1le I nte net as a meanmg
perceIve � . fuPIIeway of the extent to which users can
of being, whether completely 3
separate from or mextncably . l. tertwined with their physical lives.
The focus of research from t IS
_

� . pe rs p ective shifts away from looking at

the .Internet as a tool or a cultu:al � .ace I p a n d noves toward the ephemeral This capacity of the Internet is, for many of us,
0 f
terntory exp onng l ' the ways m dIVl d ua 1 s
, l
' n a co�puter-mediated society everyday communication with others. We can disregtaken for granted in Our
t
construc an . , d experience themse 1 ves an d others because of or throug
, , Imesh to communicate instantaneously and inexpensively wiard location and distance
Internet comm,umcat�on, Th'IS, Conce tualization crosses many dlsclp ' the distance-collapsing capacity of the Internet allows th people, Logistically,
and often studIes the mtersectlOns ofsocial identity, body, and technology to participants aro be. The researcher can inctheludrese peo
earcher to connect
(see, for exampIe, Benedikt, 1991; eherny and Weise" 1996 ' Featherstone u navailable for stuundyd ,theThglo only increases the pool of paple previously
and Burrows, 1995,. Sondhelm, ' 1996,' Stone, 1996; Turkle, 1995), also provides the potentiaisl not rtic
for cross-cultural comparioons thatipawents but
readily available previously for
_

. to consIder , the various ways in which people use and m�ke re

al and financial reasons, In a worldt no
It is essentIal where potential participants arepraonlctic
sense of the Inte�net as, a commumca ' t'JOn medium - , because sensemakmg wireless connection away, distance becy om a keyboard click and fiber optic or
P rac tices dif fer wI d ely , On e ml ' ght ake sense of it as a tool, focusing on the matic consideration in research design, Ryenes (20 almost meaningless as a prag­
abilityofthe Internet tomake informat'JOn see'kl'ng and retrieval more eff. icient
m
use email to conduct a long-term interview study02) , for example, was able to
and effective, Anoth�r might pe�ce��� ' the Internet as a place, focusmg on wit
in Tanzania from her horne location in Norway. In thi h an Asian entrepreneur
the cultural �ou� dan� c:e�:�� !n eractions rather than on the channel as an extension of the researcher's and participant' s case, the Internet served
for commumc,a�lOn, T e r t perceptions can influence greatly the Research can be designed around questions of interas bodies.
way people utIlIze and ta�k about the Internet. As well, the researcher's own unbound from the restrictions of proximity or geo ction and social behavior
perceptions will influen�e the way 1le o; she observes and interprets discourse be selected on the basis of their appropriate fit withingraphy, Participants can
in online contexts. Bemg 0
aware f he distinctions can help one better rather than their physical location or convenien the research questions
understand the context. (2000) argues that the ethnographer' ce to the researcher. Hine
s not ion of cul
reconsidered given this capacity of the Internet. tural boundary must be
traditional, geographically based means of encaps Rather than relying on
T H E I N T E R N ET AS A GLOBAL, I NSTANTANEOUS study, such as national boundaries or town limits, ulating the culture under
N ETWO R K O F I NTERACTI O N more accuracy in using discourse patterns to find ethnographers might find
, WI' �-h artici ants and collecting data, the Internet Senft's work (2003) exemplifies this reconsideratboundaries.
As a tool for conne�tmg from geographic to discursive. Senft studies the ion of cultural boundary
offers many int,erestmg PO�lbI'ffI Ies� to too does the Internet provide a means "webcam girls," a project that would hav sensemaking pra ctices of
of understan�mg be�t:r t e wa� hat language constructs and maintains e bee n hig
for many reasons, In this long-term project, Senft hly unlikely ten years ago
particular socIal realIties, Th In ernet continues to provide environmen s women display many - and sometimes all of the accesses websites wherein
within which researchers can�mteract "':1'th or gather information from partI­� the use of single or multiple video cameras in their private activities through
0
cipants, Whether ne sets up an envIron ment 1'n which to interact with video displays themselves and talks to several ir homes. Senft studies the
participants or, observes n�tura11 �c rring discourse in discussion boards, the Internet to express self or make personal and women about their use of
weblogs, real-tIme chat envuonmrl� ' �mail exchanges, and so forth, one must ofthe Internet's network of connections, participanpolitical statements, Because
consider the �undamentals of ho; ���1::�e ommunicating with one another engage in this activity, regardless of where they ts are selected because they
in these enVHonments and ho �f1uence interaction tende�cies does Senft e access to archived activities of live in the world , Not only
and outcomes, One can also exp1ore means by which to utilize creatively can sustain hav these women, but also she
certain environments to truly�ug�e�t the way we come to know the sub'Jects allows her tocon tact with these partiCipants over lon
study the way their perceptions and disgplaperysiod s of time, wh
of our. research and better un ers ar: the complexity of language and social time. of self change oveichr
realIty, Here, I examine three essentIaI aspects of Internet communication to The global potential of this medium is often con
0
consider in the development f any quanative 1 research endeavors re1ated an achievem flated with global reach,
to the Internet: made Wildly ent that relies on global access, Popular
speculative and promising predictions abomeutdiafreeaccounts have
global access
1 00

1 01
 I NTERNET COMMUNICATION AS A TOOL FOR QUALITATIVE RESEARCH

ANN ETTE N , MAR KHA M

TEXTS

many articles wri tten in the 1990s On the other hand, authenticity is questionable in any setting, online or
to the Internet e(sefOe, for exaS :p1e" the) Cur short of the
rent statistics fallicalfardevel offline, and the search for authenticity presumes not only that people have
by G�orge Gil� �ar;/y ��� iS b:�:�;: �he pace of technolog me . opm:nt real selves to be revealed, but also that the authentic reality of a person is
predicted mar . mf . d. for widespreathe d and xpe nsive revealed by the person's physical presence (see Silverman, this volume). The
far exceeds theess. . ErasventruinctuthoresereqcouUlrnetnes at the top of. 200list2)., diff usio n. of Internet appears to engender and highlight the dilemma about authenticity
high-speed accy mto , the h0me has not exceeded 6701/O (lTD, For hta-
qua for researchers. First, one's online identity need not correspond to physical
this technolog see al studies, this medeium therefore markers, If a researcher seeks to understand the physical person yet relies
tive res. ear. chers atekm.. Mo g to conduct truly,�mglob
. re genera:11ea.�ch sps eak . ing of theainissuconservof access , as. on anonymous CMC as a way of discerning this, authenticity will arise as a
remams mad equof techno y, re should rem ativ e m problem. Second, it is difficult to "read" participants online. If an interyiewer
a double edg: ns hat theloggen l l? P�{ ace accesses andandutildegizesree as thoset··
Inte rne seeks to know the participants in depth but does not spend enough time to
their expectatIOmunIc: atIo. n technoerlogl .
es m the same way get to know them and understand their idiosyncratic discursive tendencies,
mediated com rich contexts. authenticity may be considered a problematic issue.
in academicallye;c of g hic distance, the partive ticipant can remainonyanomony- Mann and Stewart (2000) take up the question of hoW researchers have
As a co.ns'equ � US a�;:;rage s for the qualita s reak esearcher. An houust approached the problematic issues of anonymity and authenticity in detail
mous, ThIS has ronm 0VIO ly wit (208-15), noting that solutions must be both pragmatic and case-specific,
interaction envi aboents may �l\OW parfIC1' �res ants to � moreons,freeThis feature Regardless of whether one believes authenticity is possible at all, research
restraints broughdIe� s ofutnsk �y sooa :o�ms� �eha�iors and onventi cce ptable design must fit well with the questions being raised,
ially una
is useful in stu n (20 3) expy lor�ese�lan the "pro-aornorsocexia" movement was
attitudes. Johnso �
:h=�ices:rkin g face-to-face with par ticipants,
born and �volved ?nl�ne. R�t eerpract crea ted. The Internet as chrono-malleable

she exammed the n dIscurSIv in websites they had ideas As well as collapsing distance, Internet technologies disrupt the traditional
infrastructur� of the Interne�;�;�!tr���nor exics to express iderntities.
thei
use of time in interaction, with several intriguing results for qualitative
t c�nsur;� rmation tonection to their actual
and values wIthou vIded s info tigmatized naturearch the rese er in f,ocus .groupsr, researchers conducting interviews or focus groups. Because Internet tech­
They may �aves,pro because 0f the ;d have been muchofmothisre eatm g disorde nologies accommodate both asynchronous and synchronous communication
or in intervIew asbuta res r wou , diff icul t; in this between individuals and groups, the use of time can be individually deter­
Johnson's taskable to accearesscheove nson, 2003) . Norwegians mined. Though an individual's choice may be somewhat limited by the
case, she was 20 2) stud ed thr 500s siteses (Joh s of specific technology used, the Internet marks a significant shift from previous
Bromseth � ? and bIs� xual� e:g � a!t��uPhracsheticecould ak
have obtained technologies for interaction, which forced simultaneity (telephone), took a
exploring l�sblanIsm-face sett:m�s,l�(�as ��ikely th�t she would have obtained long time (letters), or provided only a very limited middle ground (answering
these data m face-tdIv? erse samp e Partly because the population of Norway machines, facsimile transmission of documents),
such a rich and d : feel less anon�mo us in general This feature of Internet technologies has several pragmatic advantages
is very s�ai�;� �;���r�u�;�������e:;�sexual normati ty, the likeli.hoo� for the qualitative researcher. Complications regarding venue, commuting,
(Bromset , 'e-to-face partIcl . , pants 1' n the manner Bro mse th achIeve and scheduling conflicts are less restrictive when interactions occur on the
of involving facis 10 . Internet. As with the distance-collapsing capacity of the Internet, the elasticity
via the Internetgmatlc:--"ally, anonymitY and geographic distance ease certain of time is often taken for granted in our everyday interactions, We rely
Viewed pra . s: the part'Icipa nt" has many waytl'all s t0 wI'thdraw fromoththeer on our ability to send a message at times convenient to us, secure in the
ethical consId. eral'kItIOn
l"h1 . d of mamtamm
00
. g confl'den ' ty wis wh high. The . tici- knowledge that the recipients will access and read our messages at times
study, and. thef couerse t t.he res ear ch er doe s not kno o the par convenient to them. Beyond this convenience, Internet communication is
side of thIS, 0 t in any, ISemtha which for been a che
som e res ear rs persistent; conversations can extend over long periods of time, picking up
pant is, at leas about aut:eno�ICI �e,�, t \' bsl:u:�r'aut
y. �� hentici ty has stick· where they left off with greater ease than in face-to-face settings, where
raises con cerns d, interacting wittheh memory instead of archived text aids in the reconstruction of prior events. The
.
ing ?�mt f0: ma ny In ternet resronments results in han
ear ch ers . 0 n one
the loss of many of ability to archive accurately and trace precisely the history of conversation has
partlopants mqUan� �ymous e;nvior ranted in face-to-face interviews and been used by researchers to conduct longitudinal studies with individuals
interactional ,al1t les take �al non-verbal information may constitute (Danet,200l), to follow the development of groups over time (Bromseth,2002),
ns. ThI
observatio gap of mf S ab �ence 0 ��nYsifor the researcher who relies on it as a way and to refocus attention and discourse about certain events that otherwise
a meaningful ormatlO would degrade in the recesses of organizational memory (Baym, 2000).
of knowin g.
103
1 02
ANNETTE N, MARKHAM 0 TEXTS
INTE RNE T COM MUN ICAT ION
AS A TOO L FOR QUA lITA TIVE
RES EAR CH
"
In the midst of a c nVerS��IO�, . ynchronous or asynchronous, users
s specific software or hardware: contextual aspects
appreciate the opportun�?ty t?re ec ,O a comment or message before respond- be added to the pro s. We adapt and use technooflogbeiiesngtowisui th others must
ing and, 1'f th, e c?mmumque IS senS!tIve .n or l'mportant, to review the message ,,, whether or not thesecesuse are those intended. For example, uset our needs,
before sendmg It. In the resear�h �'eUing, these taken-for-granted capa�lhbes employ more than one coms mu rs ten d to
can significantly enh��ce bot \ ; ;coP; of a study and the collectIon of simultaneously as email is beingnicdow ation technology at once: surfing the We
aded; additionally, at any time anb
information from part1CIpant�" In 9 was conducting interviews online, Instant Message might pop up onto thenlo
it became clear that t�le qu�stIOns ask:� could be carefully considered and within a new or continued conversationscr .
een, occasioning a typed comment
On
' dunng
rewntten ' the mterVIeW, ln one I'nterview, I beg'an to write, "Would beneath the surface, social reality is both perthe surface, this is multitasking;
"

you descr�be yours�lf , as an "

h r dd' ct7 /1 I{ealizing that the outcome 0f cei ved and constituted through
this questIOn was lImIted by ��:: ���:;at' � e�ased this question and modified the interplay of time, spatiality, technology, informatio
Even in stra n, and the other.
it to read: "How wou, Id you defme
an excellent c1101ce ' IS 0f less , an
, lmpor
Internet addict?/I, Whether the latter was,
th1's discussion than the fact that It were not designighedtforwfacardilitinfateorm
to
ation transmission env
a sense of presence, progra ironments, which
' than the, flrst , wh'lCht0was
t ance into shared spaces as the meanings
by the interactions transcend the lim, itat relationships, and commumsnitcan evolve
is a better questIon
Even in a synchronous enVlfonmen ' t, I had theboth leading and close-ended,
opportunity to reconsider my are interacting. During an online focion
ies cre
s of the programs in which peoatepled
message and, reformat my query,
' are made POSSI'ble. by stopping time during me, participants used multiple technolous-giegros up Sim
discussion conducted by
Backspacmg and ed'tI mg
'
P
complicated data collection but facilitated in-depthultaneously in ways that
" mterrupt10' ns gaps
an interactIon,
connectIOn,
, auses' a! ld ' are expected m' CMC because . of speed
d the fact t1lat many users are multitaskmg, , Inof environment was a synchronous chatroom, which alloparticipation levels. The
real-time partiCipation among several people. Each perwed for pseudonymous
asynchronous me d'Ia such as ema!'I or threaded dI'Scussions, these pauses can
'm

be posted as SOon as he or she clicked the send or entson's comment would

, c

be qUIte' 1ong, perhaps even

rather than stoppm, ? ,pomts, , weeH1ernng
(S or montlls, yet can still be considered pauses
' d'IScusses this as "persistent conversa­ scrolled up the een as the conversation progresse er button. Messages
participants whoscrhad d. In one session, two
tion'" whereby partlcipants understand and, work around the disjunctive and pre vio usl y bee n act ive con trib
as actively as others were. Because of the progra rs were no t talking uto
1
fragmented structure 0f 'nteractions (1996, 1999) , mmed environment we
3

were using (Inter net Rel ay Ch at), I wa s abl e to sen

Not only is it useful to C 01�.s1'd�r th,e way that time can be utilized as a to talk privately, which, when accepted, opened a new d one of them a request
malleable construct in quaht, ,atlve mq�ury, but also it is necessary to consider only on our two desktops, in which we chatted privat screen that appeared
that as modes of inte racbon
, saturattmuecon tO merge the technologies for me that she and the other non-talkative participant hadely. The participant told
' . , mcre
communIcatIOn , asmgly e our everyday 'lives (Gergen, 1991), ,If as we were, in a private room, discussing one of the actually been chatting,
we take senous1y the co11apse 0f tIme-space ' dI'stI'nctions (Giddens, 1991 .) In in depth. group's earlier issues
the "knowledge age," these become not slm ' ply pragmatic but ontologIcal My discussio
considerations. group conversatnionwit, hexcthiepts parthaticit exc
pant was similar to whispering during a
hanges in the larger group were not
disrupted. Her private chat with anothe
fu�m� u m���� side conversation, one that added valuabler par dat
tiCipant was also an extended
unobtrusively in a physically present focus-groaupandsetcould not have occurred
Communication VIa. the �nternet occurs in multiple modes, senally ,
or must be captured and archived, which requires ting. Of course, the data
simultaneously. Whether sponsor�d by software and hardware, a person's informed enough to realize this and tell the researchethat participants be well
individual use, or the emergen:e 0 dy dic or group norms over time, these valuable information when they engage in these whispr that they are producing
multiple modes operate on th� ense:��in practices of users. Consequently, the researcher - conversations. ered and hidden from
the issue of the Internet as mu � bec�mes meaningful when designing
tl-mo
_

In anothe instance, when a participant appeared

interactions in the res�arch c�ntext. found out, usir ng to stop participating, I
In technical terms, mteractlOn can be synchronous, asynchronous, anony. by an earlier comthi s same technique, that the participan
ment made by another participant. Het had been offended
mous or non-anonymous ' One
. t'can use text, graphic images, sounds, . . andor no longer certain tha t his contributions to the conversation were des sta ted that he was
video, exclusive�y or m, .combm lOn. Pro rams can simulate letter wntmg that perhaps he should
passing notes, dIsplay mformatlOn � WI'th rew contextual features, or provide this in a private, onlinewitdishdrcusawsiofron,mI the study. By talking with himiredaboandut
a sense. of shared
. Itspace. . t'lOn VI'a the Internet involves much more offending comment was not directed at wa s able to convince him that the
In mterachona erms, commumca
. ' . t1 . valuable. Certainly, this could have haphim and that his tributions were
pened in the coucon
than accomplishing the mechamcs 0f these multiple modalities or learmng located focus group, but our private sid eline conversatrseionof defa physically
le

1 04
used the
1 05
ANNETTE N. MARKHAM TEXTS INTE RNET COM MUN ICAT ION AS
A TOO L FOR QUA LITA TIVE RESE
AR'c H

situation, eased the participant's misgivings, and allowed the larger group Jenni fer says , " Obv
i ou s ly , I have the
conversation to continue while we were sorting this out. The participant re­ I wan t to s ay to you
. . . as wel l as how
cho i c e to typ . e in wha t

entered the conversation and later told me he talked about the offending you ' " l' . e . , language cho
I want to s ay it
ice , depth of pvnlan '
atl' on , sml' 1 lng
.
to

comment with the person who wrote it, with positive results. These examples etc . "
� ...to-'.
,

illustrate how a researcher can take advantage of multi-modal features of " For e_xa.tnp 1 e , you
Internet communication. Allowing multiple conversations to happen at once, Jenni fer sugges t s
" may or may not-
h. ave
when these do not negatively affect the main group discussion, can add depth ' t 0 wh.at I say
not ed that I lns ert " ac
t i ons " In
s l lke " X exp .l aln
. , . - '. )
and texture to the discussion. thJ.ng
'
' s " be f ore launch
ing into wha t I h
, or
a e �
Whether the technology provides the multiple modes or the users adapt to say , or emphas
i s around cer tai n
words with ast eri
sks
technologies to a multi-modal way of thinking is less important than the fact '

that this multi-mod'al function powerfully influences the way users perceive
Jenni fer con tinues
, " things that I '
the c onvers ati on .
"
ve found tend to hum
ni ze �
'.cforontexts and interact with one another. For researchers, this has great potential
augmenting traditional approaches and creating previously impossible Jenni fer bel i eves
they guide bot- h whe
the l i s tener is goi
re ,she ' s gOl' ng and
where
methods of interacting with participants. Jenni fer
ng .

s ays , "I think i t ' s very

demons tra tes
Control over the communication process more
act ion be tween X#
Consider the complex combination of oral and written styles, the choice,
hel pfu l . . . I thi
, nk i t
att ent .l.on to the qua l i t y o
o f' per son s who are
f the lnt
' er-

granted by anonymous software, to create alternate identities online, and the int era c t i on . "
par t i c ipa ting in
the

ability to stop time in the middle of any interaction. These means of being in Jenni fer has always
the world with others are associated with a feeling of greater control: control ��
' on- l J.ne �
found i t hel pfu l
t 0 be very des cri
ptive
over the content and form of the message, control over the presentation of self, "
us , b / c It,
env ronment s , whe
ther synchronous
'
or asynchro -

and control over others' perceptions of the self. The issue of control warrants a f u 1 1 er '
glv es peo ple more
to work wj. th '
more r ounded sen se "
discussion, because it is an inevitable part of doing qualitative Internet-based etc .
o f your thoughts
" fee l ings env ' ronment ,
l
research.
Internet communication can provide the researcher and participant with i rself in:he third p�rson, Jennifer uses a variety
the opportunity to reflect on and revise their statements before actually of ;:e����� �; �i1?;:�i�� ofac��le�
uttering them. Most participants interviewed by me online believe that the . ructivee ':vIn sev
certainty. Th'IS exc,erpt IS. Inst at she belIeves to be conversational
ability to edit affords a higher degree of control over the meaning of how careful attentlOn to the form of the texteral can
ways, Firs" t it allows us t �ee
hel 0

the message and the presentation of self. Whether or not the producer of the os
message can actually control the presentation of the self through careful ����� ;:���:�ai, :�:e�ctions een b
Even i� one does no� ;;a:��: ��:f�����t���
' serv 0
editing is not as relevant as the faith placed in editing. structure of sentences can remind thatlO e
ns Jennifer ma kes about the written
rese archer that understanding between
Jennifer, a participant in an online interview, trusts that careful attention interviewer and " n etwee
to the construction of her words will give her a higher degree of control over achievement that�;���;�: :i����y care�u�artici atte �antn toand participant) is an
ntIo form of the message
the conversation. When asked how she tends to interact online, Jennifer replies as well as content.4
Second, this erpt gives us insight into the w�y people mIg'
(the programmed environment in use displays the responses as follows): Internet commuexc nic atio n Som e w'n pay 1
ht per ceIv
1?se attention to the use of languagee
,
in this medium, Others' ':vIII pay very clIttl
Jenni fer s ays , " I would say that I become very attuned to
To �:::' d: ��ntr�:� her�IS a tYPical excer et from
'
attention to form or content
*what * is being said and *how* it is being said - particularly
the t y , h, w spen ds up to fIfte
0
� en a different participant i�
in a synchronous conversation and l i kewi s e at tuned to howl
community, is responding to the question' "Why dhouyours spe a da� In' her onl'Ine
am 0 n so much time
what I s aying as part of that conversation . "
online?" We pick up the conversa t'lOn ml'd-stream:
Jenni fer says , " I find mys e l f thinking a lot about what is
the " right " thing to s ay . . . trying to make sense of and �eth says ,
" yes but I think I
l ike it this way b
J us t ecau s e I can
interpret the mood / a t t i tude in addi t i on to the words , such typ Wh t
ornmes to mind and
� not have to think
am it as mU h t l' about

::: :�:�
that I can be sens i t ive and focused in what I saying in � � to be co
my finger s then mmuni cated bett er through
reply . " o

1 06
1 07
r

ANNETTE N. MARKHAM * TEXTS

INTE RNE T COM MUN ICAT ION
AS A TOO L FOR QUA LITA TIVE
RES EARCH

Beth says ,
vo ice my ; m "
"I can type what i ' m feel ing better then I can " I can wri te who
I t
- rul y am
c l ear ly and dir ect
ly by edi t ing . "
fl
I
can edI' t the tex
t and con tro l how
Beth s ays , " f eel ings it just c omes a little easier s eeing I pre sent mys e l f
. "
things to answer then hearing and having to answer I l i ke " Through edi t ing
, I have a l o t mor
e con tro l how oth
per ce ive me . " ers
to worrk with my hands a lot "

Beth says , " i t ' s j us t what your typing that count s " " I can be anythi
ng and anyone I wan
t to be in the tex
t. "
Beth s ays , " this is a place where you can get to the real
i l e e f
person and not have to overcome the obs t i c l e of looks and
� : :: :
o� �!t �d� : ��:��t�� � �: �a��ot��on::����y :�� s m �pJe interpreta-
:� t:
having people j udge you by your appearance insteafd of the
!�iS set of responses, respondents indicate that they (the ;r�;e:'/�:���r�ther �ro7
I
' .
real you here your real self comes ins ide the things your
e message, thereby controlling the outcome . con
wrte "
tells us these users perceive that the h v . As a group, thIS" set of responses
way they are perceived by others be !au � � h' Igh degree of co�trol over the
Like Jennifer, Beth believes she can better control how others perceive her
By contrast, consider the statements beI :Y con trol the"Outgomg message .
in this medium. In Beth's case, the control is not in the form of the message . ow , utt ere d by the exact same
but in the meaning "inside the things your wrte [sic] ." In contrast to Jennifer's par t'lClp
' ants m response to the que'stion . "What
' are som e 0f the lIm
' itations
typical writing style, Beth's grammar and spelling are, by any standard, of this medium for you?"
terrible and most likely not deliberate. These errors can be a result of typing
" In this medium ,
fast, not editing the text, or being unaware of the errors. Regardless, the nobody knows who
I rea l ly am . "
example demonstrates an important point for researchers: discursive practices " I can ' t t e l l who
oth er people rea lly
are i f I j ust hav
in this medium are wildly different; form can be unnoticeable or glaring; tex ts . " e the ir
and content cannot be disconnected from form if the form is glaringly dis­
" I t ' s di f f i cuI t
junctive from traditional writing norms. The researcher cannot help but to know the rea
l i ty o f s omebody
if their
wrl. t lng
' doe sn ' t a f f e c t you
be influenced by the form of the message, which in turn influences the or speak to you .
"
interpretation of the meaning of Beth's words. Even as Beth believes that she " I t lS
. a game ; eve rybody wears
masks . "
communicates better in this form than through her voice and that people
reading her words will see through the form to the real self, the question arises: The contradictions in these res� ' are .
the benefit of the medium is tha� o�ses cunous. PartICIp . . ants indicate tha
"How well does she represent her self with this use of language?" These
examples are good reminders that participants are likely to have different �� t
of self to the other in the interactic: �con�eys an accu.rate or desired sense
t
habits, skill levels, and experience using Internet communication. The same At the same ti e, however, they also md . lcat on y message IS the message sent).
might be said of spoken language, of course, but not with this same degree the text canno�c��vey an acc rate or rea e that one of the limitations is that
of difference among speakers one would typically classify in the same category interaction (the y messae g e�lS the mes l sense of the other to the self .in the
sage recei d) Wh h t '
(in this case, native speakers of English, at least high-school educated, self­
described heavy users of CMC). The Internet intensifies these issues for
:�:��:�::�s �fer�: t �n essentially
!; self-ce�;er�d po:�ti:n :���!a�
: l
t se
researchers. researchers u;in the Intern t en :onstrates at leas.t .two considerations for
o n e ct �. t�1 pa�tlClp
The text is a fundamentally different space of observation and interaction still adjusting to 1nternet me�i� a�� �:ve � . ants. Fi�stl people are
than sitting next to the participant or observing interactions in natural settings. ways of performin the 'self throug 1 t� IS m hve nd
� � pOSSIbly unknown
Careful reflection is necessary to make sense of how we researchers are individuals use Int�rnet me d'l� m . SIm� ese me dl � . It IS hasty to presume all
engaging in or observing these interactions. In most cases, it is recommended two people using the s�me me lU . � ma I ar :-vay ;
� mformation collected from
to treat each case individually and apply appropriate standards, practices, and of the way they perceive th . y Y Ield mcomparable results because
procedures to each. Even so, it is impossible to predict how individual
participants define, use, and respond to specific computer-mediated media
�
bec�use it falls outside the re:e:�h������!t�� �� ��at ca� go �nnotic� d
n n
e ng these n:edia, we are�i�:ly�� ::�:nt��:
and contexts. Take for example this series of benefits about synchronous,
anonymous, text-based CMC, cited by several users participating in a study.
�:!��;�:;���� :������: ���
Does the res earcher believe t S �tee��� nameles� and faceless settings .
m

These benefits are mentioned in response to the question: "Why do you like Conversely, does the researchethat r bel
y mess�ge IS the me ssa ge sent?
using this medium to interact with others?" message received?. On the surface, ieve the only Important message the
th ese may seem simple questionsiswit h
1 08
1 09
ANNETTE N . MARKHA M TEXTS
straightforward answers, but even with careful reflection, it is easy to beli�ve
that our own utterances are clear and unproblematic. It is valuable pract� ce,
whether working in Internet settings or not, to engage in �ritical self-.reflectlOn
about how questions are being asked, what presumptl0r:s are b�mg made
when observing focus groups, and how our own preconceIved. nobon� of the
communication process shape our interpretation of everyday mteractIOns.
Push versus pull modes
Anyone who markets products on the �eb or wh.o teaches c?urses onlme
.
can verify the importance of using the nght medIa for �he nght pu�p �se.
Push/ pull considerations are vital to whether or not the mtended reClpient
notices or attends to the message. Push describes a technology that pushes
the information to an individual's computer or handheld device. This term
also refers to the extent to which users feel as though the message is pushed
toward them, requiring attention to read, trash, file, or otherwise do. sorr:ethi�g
with. Email is a good example of push technology; messag�s arnve .m a 11st
and, putting filtering programs aside for the moment, r�qUlre attenhon .and
action. Other technologies can collect and push news Items from vanous
sources to one's desktop, text messages and weather reports to one's mobile
phone, or flight schedule delays to one's P�A.
. .
Pull technologies require a more proactIve approach by the user. the ldea
of the information is so interesting, important, or intriguing that the user
will be compelled to seek out, find, and attend to the message. Although
distinctions between push and pull technologies are becoming more and more
blurred by the evolution of various media and usage patterns, the concept
is useful as an initial categorization tool for the researcher. .
The following example from a teaching experience illustrates the �mp.or­
tance of push/ pull considerations in designing an effective commumcat�on
environment for active group participation (focus group) and collaborahve
learning. In a recent course focused on hypertext theory and des�gn, students
were required in one assignment to redesign the Internet-medlated aspects
of the course. I intentionally designed the course to overwhelm the s�ud�nts
initially with multiple media choices and requirements for commumcah�n.
Each week, students were required to log into a password-protected webSIte
where they could find links to the syllabus, schedule, and announcement�.
They were also required to use a threaded discussion board accessed from thIS
site. In addition to this Web-based system, they were expected to check their
university-assigned email account, where I sent them. bot� ind�vidua.l and
group (listserv) messages. Many students did not use theIr. um�ersIt:-assigned
email account, which meant they had to begin to check theIr umversity accounth
or notify me to change the default address I used to ser:d email. Finally, eac,
student was required to set up a blog (Web-based J ournal) to post then
responses to readings and other thoughts related to. class. !�ley wer� expected
to read and link to other students' online journals m addItIon to mme. the
As planned , too many different modes of comrr:unicatio� vied . for
students' attention in this configuration, and they qUlckly realIzed tlns prob-
1 10
INTERNET COMMUNICATION AS A TOOL FOR QUALITATIVE RESEARCH
lem. The remain�er of the course involved solving the problem to meet
the goal of bu�. ldmg and maintaining a productive learning community.
Stu� ents examm�dj among other things, the push versus pull aspects of
�anous c�mmU�Ication choices. When some students suggested, "Let's get
nd of emaIl and Just use the password-protected website for all information "
others r�sponded that this would require unsolicited attention to the cours�,
a proactIve appr�� ch .that could not be presumed. In addition, students aptly
noted a key usabIlIty ISsue: before the student could even view the front page
o.f th� course, the p�ssword-protected site required three keyboard events,
SIX c�lcks, a �d effectlve navigation through three screens of informafion. A
publIc We?SI :� with easily accesse� information would be easier, one bright
student saId, but then ,:e would shll need to remember to actually go there."
Aft�r several weeks of hvely debate over various issues, the students finally
deClded on the following elements:
.. Public web page, which all students agreed to keep as their browser's
� efault �lOme page for the duration of the semester. Most relevant
mfor�ahon appeared on this first page.
Runnmg chat board on the web page for general student conversation
and studen: anno.uncements (similar to Instant-M essaging software).
I!>
.. Threaded dISCussIOn on the front page of the site for more serious' lengthy,
course-content-related discussions.
• L�nks to all student web journals for those who are interested.
• LInks to the course syllabus and schedule (transformed from documents
to HTML documents for speed of transmission and ease of reading).
�n addition to this single web page serving as the course site, students
belIeved that �oth th � listserv and email should remain active. To a person,
they hypothesIzed (nghtly or wrongly) that any information sent from the
profess.or to the student is vital and should be pushed into the student's
ImmedIate awareness.
This extended discussion of a single assignment in an academic course
und.erscores the considerations that go into the design of a communication
envIronment. . In parallel fashion, research environments utilizing various
Internet. medla .rr:ust undergo similar evaluation, because each decision
conc�rnmg p �rhClpant communication makes a difference. Testing various
medIated er:�lronments can help one discern which is most suitable for the
type �f parhc�p �nt. Collecting life histories via email may be satisfactory, but
allOWIng pa�bClpa.nts to create ongoing life history accounts on websites that
they can �eslgn WIth color and images may yield richly textured results. For
an �nterview study, real-time chat� ooms may provide anonymous partici­
pa.hon and spontaneous conversatIOn, but email interviews may be better
sU1:ed to participants who have busy schedules and desire time to consider
theIr responses. The key is making a conscious and measured effort to match
:he r.node to .the context, the use�' s preferences, and the research question. If
ne IS studYIng naturally occurnng data, this issue may not be salient to the
111
ANNETTE N. MARKHAM TEXTS rj
INTE RNE T COM MUN ICAT ION
'J

AS A TOO L FOR QUA LITA TIVE

RES EAR C H
process of collecting data, but because push/pull variables influence inter­ this confluence of technologies and
actions within the contexts under study, knowledge of the possibilities texts operate is crucial. sub)eC
' hv" lty, clo se attention to the
way
and limitations inherent in CMC design can aid in the process of analysis and The Internet is not novel in that ind ' .
interpretation. groups, and technical capacities constit dua1 use, habitual pr�ctice across
��� patterns of temporal mteractions,
The overriding message throughout this discussion is that reflection and
building social structures th t b :��
cesses describe any langua : s ::?m :; concre�e re�lities. Th ese pro­
adaptation are necessary as one integrates Internet communication tech­ enabling us to view these pl (�ces�s �f ' . nternet �s Ul1lque, however, in
nologies into qualitative research design. Adapting to the Internet is one level primarily textual interactio'ns 'w t socIal �o�struchon as solely dis cursive,
of reflexivity; as we use new media for communication, the interactional conduits for the transmissio�l 0: :��� m s us ��at texts are mo re than
challenges and opportunities can teach us about how to use these methods. ceptualization of texts: :g. I-Ie cntIques a common con-
Adjusting to the individual is another level; as in face-to-face contexts, a
skilled researcher will pay close attention to participant conceptualization Texts are placed in service of
and utilization of the medium for communication . Without having access to the examination of ' ot her' , sep
phenomenon. From this standpo arately conceived
int the tex; p ur�ortedIy com pns
physically embodied non-verbal features of interaction, the researcher may accessing these phenomena- '
phenomena eXIs .' es a resource for
want to deliberately address these concerns with the participants so they where the tex t operates as an .
tmg 'beyon d' t1le text, as It
. essen t'IaIIy unexammed. were
, wm " d of neutral'
may aid in the interpretation of discourse. Alternately, the researcher may dow' or 'channel' to them. (199
7: 81)
cond '
lil t, a Klll

want to adjust his or her expectations of these possibly unfamiliar environ·

ments. If researchers cannot adjust to the particular features and capacities Referri g to Rose's
of Internet technologies, they may miss the opportunity to understand these �h��;��e�����xt(9 60) notion of the world as " d
��edia�e SOCia1 interac tion an� b:��
. .
"
phenomena as they operate in context. As Gergen (1991) notes: if we are to y be m t0ta l agreement as t? what com . s:� �;� n�:::�:'
i:7 �
in conSidering the centrality of texbrs H�r�:
survive, flexible adaptation and improvization will become our norm. Watson's point (a point made also . tl� pnses text,
Along these same lines, Carvajal (2001) reminds us that any decision made 't ge m IS �olume) is well taken
about method should derive from a conceptual and epistemological level of meaning. As a context almost eIIt'Ire i e n.egohatlOn and construction
rather than from a procedural level. In discussing computer-assisted quali­ exciting location of social meanm . g anI 1 comP:ISe? of text, the Internet is an
tative data analysis software (CAQDAS) training issues, Carvajal stresses that as discursive practice) . One can a1so orgal1lzabon (de fini, ng text broadly
anyone using computer-assisted programs to analyze data should incorporate Foucault's approach to the stU dy 0f c usefully reca11 Pnor ' s d'Iscussion of
"critical thinking instead of mechanical thinking" (section 3.2). "To know a becomes the focus of qua. litatI've explora ultUre (thI' S volume), whereby the tex
. sub .
tIon rather than the aIways-elUSIv .
t
software is to know about the methodological implications its use has for "knowmg ject." e
qualitative methodology" (section 3.2). This thinking applies also to the use
of the Internet in gathering information for analysis.
Internet technologies all ';!��1I" a :�
construction process in a v�; v ve researche:s to study the social
or minimize the visible producYts of m a� ' Because It can constrain, hide,
mannerisms, and geographically base tIOn ' t era c ' cIoth'mg, accent,
(read ·. bod Ies,
. 1 stru
T H E I NT E R N ET AS SCENE OF SOCIAL CONSTRUCTION . .
focus speCIfIcally on the bu1'ldmg d SOC Ia ctures) the Int ernet a1lows
interaction. ' bioc ks of culture at' the basic lev
el of
The Internet is a network of computers that allows us to create networks of Kendall (2002), for example s en t
connection. At a basic level, one can study the connections themselves, or graphic study of an online Mu6 sever�l years conducting an ethno-
one can use the connections to conduct studies. However, remaining at this the discursive foundations and nego�?���l1lty called �lueSky, analyzing
simplistic level of binary distinction obscures the complexity with which the consequence of studying th . softwa la e �atures of thIS com . d'mu . nit y. As a
re settmgs as w e 11 as m
. ' ndall is able t0 rnal(e mte
cultural and technological aspects of the Internet are interwoven in con­ ver sat ion s Ke
e

. argumen Ivld ual con-

IS performed . In M
structing possibilities for being with others in everyday social life. Although . ' restmg ts a b out h ow gen der
UD enVIronments for examp1e, gend
we might consider the Internet merely a conduit for information transmission, ma kes by set tin g a com ' er '
IS a cho ice one
l. 11cluding male (he his) fmand1 . One can choose from a van. ety 0 f genders
the content and resultant social contexts of these networks and transmis­
sions are also fruitfully conceptualized as meaningful phenomena themselves. �
We (we, our), and �o fo�t�n�f:e Sth:' her), neu er Cit, its?, spivak (e, er), royal
;
Through the deceptively simple process of exchanging messages, complex duced for other members ' show e user rna es a chOIce, certain texts pro-
and transformative understandings of self identity, other, and reality are some members, of th1's onl'me the cor. respond mg ' pronouns. In BlueSky
negotiated. As more and more people mediate their social interactions in purposes, so that, for example, wh com mu l1lty u se th e gen d er settm
en someone asks the system whgsatfor
' joke'
gender
112
1 13
ANNETTE N. MARKHAM TEXTS
INTE RNE T COM MUN ICAT ION
AS A TOO L FOR QUA LITA TIVE
RES EA RCH
the other participant is, the preprogrammed response might read, "Not lately," actual study, the rules, practices, and
or "No, thank you" (35). BlueSky members might make gender a relevant may be distinct from or qUite similarout co�es °f mt ' eractJ" on m online contexts
0 face-to- face contex
feature of their persona or not, and those whose gender is typical (male or 1 t ts .
At van' ous evels, some more conscious
th� n ot-h ers , peo ,
, ply ple mteracting in
female, in this case) tend to be responded to in correspondence with their
chosen gender, rather than with their embodied gender. In other words,
computer-mediated contexts negotiate,
the identities and social rea lities of the r: th er t a n SIm obs erv e or cover,
if Mike (offline) has a persona named Susan (online) and declares that Susan Wheth, er interpreting naturally occ � hers WIt, h whom they are intdis eracting ,
is female, people will tend to interact with Susan as a female, even if they , rvlew, researchers
mte urnng textse or participa t'mg m ' an online
can find thi
know Mike is a male playing the character of Susan and even if they have
met Mike offline. one may find that some of the t ypI. ca t IS pomt m
�
unfamiliar territory for most resea�c���'tse �ht �engI. �1g, pe�haps because it is
' tl e. For example,
apprlcable m . the fragmented l ruIes of conversatl� d 0 no t. seem
If we accept the basic premise that reality is socially constructed through
language, the Internet allows us to study this social construction in progress . US humor is extremstru.cture 0f onl'me conVersatIOnIOn . Sar
, . y,
cas'm, Iron
or non-obVIO
��e;n��t�e :ext. Additi�nal1y,
.

e I d' ff ,
many paratextual elements are lffi� � U I t d' I
as a real, enacted process rather than a theoretical premise. Internet tech­
�
data or to categorize effectively Th e f0�11toe t�er , g e as non-meamngful
an online interview conducted b m ;mg
nologies allow the researcher to see the visible artifacts of this negotiation
process in forms divorced from both the source and the intended or actual ' mterview �xcerpt, taken from
audience. Websites and website archives, for example, can give researchers interpretation in the unembodied text. Y e, em onstrates the elusion of clear
a means of studying the way social realities are displayed or how these might
<Anne t t e >
be negotiated over time. In Kendall's ethnographic study (2002), extensive Tel l me about your mo st memora ble
onl ine , exp er ience
archiving of interactions gave her an immense and enormously rich set of data
to work with. i don ' t know ,
<sh eri e> gee ,
s o many , some are per son
Of course, multiple variables influence the way we make sense of the world aren ' t . al , some
and this confronts researchers when making decisions about how to approach <Anne tte > great
!
the field. In designing the interface with participants, interacting with
' tal k a 1 1 you want '
<Annett e> cho ose
anY - - a l l
participants, and analyzing human expressions and experiences in naturally '
<sh eri e> we l l ,
occurring settings, researchers will naturally make assumptions about how mo s t seem to h
ave somethlng
.

to do wi th the
communi ty i bel
the communication process works, taking certain invisible features of ong to ' everyth
lng
' from pe rs onal rel
ati onships
to flesh me ets t
interaction for granted, whether or not this is warranted. Certain researchers o flame wars , ,
.
may naturally rely on non-verbals as well as the content of talk in analyzing Interviewing via CMC re es patlen
both the content and structure of conversation and may unconsciously use tendencies, and pacin of the�uir ' ce and careful attention to the ski
lls,
socio-economic markers derived from participants' clothing, accent, posture, at the end of Sherie' s �st statem���a� (�a��ham, forthcoming) , The ellipsis
es d nt
and other physical features. These are just two factors influencing the way would continue her thought m �ve m I�ated to me that the participant
researchers perceive or interpret subjects, particularly in research that relies '
prevent the interruption of' a personacc or ance WIth a rule I had devised to hel
on researcher interaction with participants, such as case study, focus group, , , , s story wh p
unavaIlable. Outlined for the par t'IClp , ant at the en non-verba I SIg ' nals are
interview, or ethnography. Warranted or not, we use physically embodied t h'IS strategy wa s useful to " atlobeginnmg' 0f each mte ' rview,
features and behaviors to make categorical assessments of conversational m
' d'
end of a conversational turn In th' Ica t e con tlnu n and - bY I't S absen ce - the
partners, which in turn sponsors the creation of a framework for interaction. because even though she used the ��ltas,e, how��er, the rule did not work,
Researchers trained in analytical methods which do not rely on visual or verbal she had completed her thought: PSIS, Shene s next statement indicates
contact with participants may be less inclined to do this, but a priori assess­
ments based on typical/traditional gendered, ethnic, and socio-economic <sheri e > are you
th ere ?
categories remain a problematic feature of social research. <Anne tte > oh ! yes , I ' m here .
These statements are not unfamiliar to anyone who pays attention to human <Anne tte > I ' m
sor ry , I though
interaction. What may be less familiar is the extent to which Internet tech­ a tendency to
t you were thi nki
ng , , . . I have
ask que s t i ons too
nologies bring into relief and problematize these working assumptions, pe ople ,
qui ckl y and alw
' ays int er rupt
Attending to these basic processes of communication not only constitutes
healthy practice for social research in general, but also is essential in devel­ �o response was forthc ing from
oping effective, rigorous, and reflexive research practices in Internet-related gIven the question I hadom , was surpri
Sherie at this pOI'nt, whIch
studies, Depending on any number of factors only discovered during the
J'us t
I tried another question to pro as k e d con ' g memorable experiences onlsin
cerm� g
ine .
mpt a narratIve account:
1 14
115
 ?

'
INTERNET COMMUNICATION AS A TOOL FOR QUALITATIVE RESEAR CH

ANN ETTE N. MA RKH AM

" TEXTS

what would design and/ or analysis allows the research project . to accommodate varying
expe ri en ce ra
ndomly ,
perceptions.
e vaIue and IS,. by ItS. nature,
you pi cked an
<An ne tt e> i f
you tel l me
abou t i t ? T�e id.ea that Internet communication has litH .
m Sherie, I chang ed tactics: fleetmg, IS made possible by habituated pr�ctlces as well as the technology.
After another long pause with no response fro Consider. message lengtlr. when transmlSSIOn
band. wIdth limitations and storage
' . . rates were s1ower because of
capacItIes on ser. vers, short messages" ,
it to o hard to
pick j us t one
experience to
ta lk
i on . . . e
p.artlcularly in synchronous ,nVIronments. ' n�cessary. In email, because
were
<An ne tt e> is f fe re nt di re ct
abou t ? if yo u want , we ca
n go in a di
the technology did not allow an�th·l�. ! b���nd pIam text �nd single spacing,
sin:ple and short messages wer m l e y to. be read. Though these. limi-
< sheri e> ok .

caj ole, prod, or ficcompel tatIOns are being overcome' short messages remam. .the norm, POSSI'bly because
ou t thi s int erv iew , it wa s dif fic ult to ult to the technology evolved in this wa ; d e �bIt IS now � social �orm.
Through more than primarily monosyllabic responses. It is dif ask ed, Take the issue of informality: il£. r:a��y ay be a chOIce but IS also quite
er
Sherie to utt ether Sherie did not like the way the questions were be migraine ing often a necessity. Simply p�/t t; :�:; 1( � lo ger than talking and errors in
ascertain t interested in the topic, she was multitasking, she had n-useful,
wh a typing are frequent even f t1 ! . s�1ll e� typists.. For the a verage user
she was no else. Perhaps this can be simply dismissed as a now people of CMC, a smooth, flowing . . conversatlon may be conSId ered a good tradeoff
or something the oth er ha nd , sin ce the stu dy wa s focused on ces ho
ough for simplified phrases �spell'l�g or grammar errors, a nd unedited messages.
interview mselves online and make sense of their experien thr
. On 0 t
expres s the iew ha s me an ing an d can no y dis mi
t be immediatelfact that the ssed. . Consider. the ephe eral ature f computer-mediated. "exts. Messages
m a bu1letm board system are often com P ared to I osHt notes or notes left on
'

langu ag e, the int erv ulty lay in thetion. Questionres doors and counters' for one's flatmate to notice. Wh. en the message
J

Re fle cti ng on tha t int erv iew , I fou nd the dif fic ua refrigerator
erbal cues to gucouideldthe interpretationbaofsedtheonsitem .IS sent, It' seems to disappear' even as much as we' know" 1·t does not. In these
were no non-v dir ect ion no t be mo dif ied bodied signaci­ls. and .many other computer-me d'1ated contexts' the n tIOn 0 0 f the throwaway
tio na l
or con another layer of complexity, in an earlier session, thielo
ve rsa s parti text IS apparent.
To ad d ye t rse lf be tte r in text because." The text quabent
she wa s , l communiques, it would
nt ha d wr itte n tha t she lik ed he ove In this context of short informal d.
pa
and that she erifelet,' s "m ore beautifult the as text than asproflecesshs and seems to belie seem likely that users wo�ld consis;e��y :!e::��� . . � s temporary and casuaL
represents Sh t becaustysele her throughou s are interview from standard notions However, users simultaneousl� or alte:nately pnvilege the text, giving it a
her statemen response a far cry state of considerable concreten ss' and Importance. This is partially because
of eloquence. munication gives qualitative researchers an intriguing optuapollyr-. any information
. via thetransmitted v'1a the Internet may be arc h'lVed somewhere.
Internet com ss the social construction of reality as it occurs tex bally InteractIOns Internet can be perceIved ' as hav'. mg a 1ong-1asting shape
tunity to wisnitne or effect, which may result in the r:articipant feelmg ' lIke he or she is on a public
This short texpp et of Sherie' s con versation is oneopamhoonnygofminelliotwnsorkofs.gloIn this stage as much as it may r esu
. in an informal conversation. It m t 11 e more commonl y bel'leved feeling of
accessible e, ts,whallerevytheingresforearattcheenrtiointnervin a cac rticipant, identity and bemg
s the pa subtle and intrigu­
specific casnegotiable during these online iew int era ctions in . Students often bemoan this ver ca ac ' �y Of the Internet. Low participation
online discussion groups dur!ng fIe1e �lrst few weeks of any school term
reality are th the participant and researcher send messages thaterdismaplakeys may be associated with fears of en anent effects: ideas spoken may not be
111

ing ways. Boplay into the construction of the context. The research in mind . erased and will likely be arch'lve� an�. use� later against the student. Second'
identity and responds wiinththethesam se judgmentse structure "
judgments ofntthemapakesrticsenipasentofand cher e way. Th inter­ the only things that should appear m wntten form for
are good or at least well-developed l l.1eas. N ot only do students tend
pu bI lC consumption
The participaon is an ongoing acctheomrespliearshm en t, dra wn fro m previousof) the to fear that speakers are held accOunt b e everything they utter, but also
of interacti sustained (or adjusted) by adherence to (or the absence they believe that they should be �ert.am rtheir statements before making
0
actions and versation. them,. since they will be wr'tt 1 en m stone. The pra gma t·IC" outco e of this
rules of con er of complexity involves the way users perceive the naturhase tyof SI'tuatlon is that ideas are less I1" keIy t be tested
0
. machieve
Another lay et sponsors a casual communication style. It is, however, ual greater sense of self-efficacy and learn to mmimIze
until t
. . , or demystify the authoritya
par IClpants
text. The Internt because of this, users conceptualize text in a similarly cast as a of the written text.
to assume thaeed, users frequently conceptualize and respond to the tex h Taking this idea to a broader cultural. a.nd h'1stoncal . scale, we can see that
manner. Indmal, lasting vessel for truth (Markham, 2000). This is true foralbot
for
concrete, and researchers, making this an issue that requires research cri tic se lf­ the tendency to give Internet commumcatIOn
is in no small measure related to the trad'IlIOn . fm' ormal and flxed characteristics
most cultures to hold written
participants careful planning to resolve. Attention to this factor in
reflection and 117

1 16
 �
INTERNET COMMUNICATION AS A TOOL FOR QUAlITATIVE RESEA CH
ANN ETTE N . MAR KHA M
TEXTS

. mg . on. gm . a1 documents a near-sacred status . Tearing T H E D O U B L E EDGE OF TEC H N O LO G Y

h reg ard , glV g . the
texts in hig d·f1 f·1CUIt to imagine as destroyin
0
. ted States °r the Magna Carta. We preserve gm�1
e o t f a t e xtb ook is alm ost as on . Social theorists and science fiction writers alike warn us that every technology
a pag. �
of the Um has a double edge and unforeseen effects. McLuhan (1962) argued that
co nstlt utIo
s
n . h ermetically sea1 ed cont am
m · ers . We tend to believe what .mg m .IS every communication medium extends the capacity of one or more of our
docum ent . I th United States, witnesses tes tify
written more than what IS hteathrde�:r ��n�s n the Bible and swear they will tell cognitive sensibilities. Writing implements and the printing press extended
the judicial system often puy a so emn�Ize their testimony. These are just our memory. Radio makes our ears bigger; television allows our eyes to see
the truth in order to verif pn�vl. 1ege texts. 1n this c�ntext' the Internet falls events around the world. The Internet allows us to connect personally and
0
exa mp 1 es f ho w we WIt . h the instantly with countless people around the globe. Wireless technologieEi allow
a few
here between text and non tex t, and we are still struggling us of it us to attach technologies to our bodies in much the same manner as physical
somew .
SIo ns th1· S cre ate s wh eth er res ear c hers and participants are conscio prostheses. Yet, for each extension there is something removed, dismantled,
ten '
or not. ded because or constrained. Postman (1985) argues convincingly that as television becomes
A student's fear ofbemg ]ud gedope
. . by his or. her texts is we ll grounoth er factors more and more prominent in our everyday lives, our attention span decreases,
n t a p ers on ma kes rates m con]·unction with com so that Americans, for example, have an active attention span of approximately
an y com me
ntI . ty and merit . Likewise, it is not un . mon to,
to rep res en t a P ers on 's' ide
. atlv. e stud·les 0n the basis of their texts; onhne, one s twenty minutes, the average length of the typical sitcom. The premise of this
judge parf1cIpa · nts . tm of
. g abIlIty
qu. aht . IS. as much a social marker as one's accent, b0 dy argument is compelling. Few of us in Western cultures can imagine reciting
typ ing an d wn · ed to know better' textual marke rs Homer from memory or attending to and analyzing oral arguments for many
type, or sk·m co1or. Even tho ug h we are tram
. of parfIc;;� · nts A reater apprec iati on f or h ow hours at a time, as early Americans did during the presidential debates
influence our interpretatlOn tex can . e p r�sea�chers make better ana
lytical between Abraham Lincoln and Steven Douglas.
users perceive the nature of ts The sensibilities afforded or limited by the Internet remain unpredictable.
decisions. ors on our As a tool of research it offers many intriguing possibilities; the temptation
Th e Internet hIghlIghts the
. . influence of non-verbal behavi to insert these as easy solutions to the problems of social research is great.
retatlOn 0f 0thers . It also illustra tes centrality
. the
under stan d·mg and int erp ating outcome As Mann and Stewart (2000) emphasize, it is vital to consider judiciously how
of the text in negotlatmg a�dat�lOn
. . onstructing reality. A fascinfoc us on basic the tool fits the research question and the context, returning always to the core
has bee n the revival of foc
of Internet-based commumc . great potential in thi s shift of us. When considerations guiding solid, rigorous, systematic, and, above all, deliberate
sensemak·mg processes . Thereme . ISs the boundaries of the study' S par ameters, qualitative inquiry.
geograp11y no longer determstram . ed bY the structure space and tim . e WIt. h· nm
the researcher �an be less con .
"

SOC . 1 rea 1 y
Ia
as it constru cts
which interactlOns occur. .Ob- se0brvI �� In ternet suse to online groups is a straight- SUMMARY
ac ces
and archiving the interactions 0f th
mg
can be accomplished. easIly' 1 oatam ese
forward process, as IS down d·mg This chapter outlines several theoretical and pragmatic issues associated with
groups. le, although the use of the Internet in qualitative research. Placed within the fast-growing
At the same hme, severa 1 ps
. et h·1cal concerns arise. For exaersmpma and swiftly shifting arena of Internet research, this chapter provides general
p ea t0 b Public memb y perceive
many online discussion grou (Fraan�(er and �1. g 1 999 . Sharf 1999) and can categories for considering both the enabling and constraining aspects of new
ction to be privat� resear��e�s (Br�mseth' 2002).
their intera � Other communication technologies, from which the reader can develop his or her
. an ger ed . by mt ru d mg
be surpnsed or
com mu n . � tl . �� ��
. ubl ic but none the less do not want own unique approach. Adopting the Internet as a means of augmenting tra­
, , u . �nd Bruckman, 200ost
groups know the i �. � 2). Ad ditionally, ditional studies requires attention to the creative possibilities as well as to the
to be studied (GaJ]ala: :002nts talk m these rou s is alm 200 impossible to foundations of qualitative inquiry, so that one's decisions to use the Internet
confidentiality of part1Clpa n 0f search e�gin�s (Mann' s2). Ongoing are both epistemologically and methodologically sound. To review some of
pres erve WI·th the sophisticatio ou t e thi cal pro ble ms and gui del ine can prov1·de the important considerations:
ent s a b
discussions and
the researche r
sta
wi
tem
th use ful� ba c � r � u � d
s �
. nformation on how others have
es (fo r good overviews, see Franketsl • The Internet is defined variously as a communication medium, a global
lt Wi th t h ese n
stewar�, !o�O.' and the ongoing ethica
approached an d dea l statemen network of connections, and a scene of social construction.
and S·Iang: 1999'· Ma.nn and th ch ers , 2002) . The shape and nature of Internet communication is defined in context,
by the ethICS commIttee 0f e Ass
ociation of Internet Res ear •

negotiated by users that may adapt hardware and software to suit their
individual or community needs.
119
118
ANNETTE N. MARKHAM TEXTS
INT ERN ET COM MUN ICA TION
AS A TOO L FOR QUA LITA TIVE
RES EA RCH
., Internet communication affords qualitative researchers creative potential talk. ObViously, we do ma ny
mo re ski lful things" 111 conver
because of its geographic dispersion, multi-modality, and chrono- explain to a researcher in .
an 111 . .
sat
. IOn than we could ever
'

\. ' •/
. t erVlew Close exa�111
malleability. .
J
1'11um111a te the
building blocks of both 111 : � t'l�n 0f texts can help
' d IVl' dua' l and 111stItu
., The researcher's own conceptualizatlon of the Internet will influence how
. .

tIOnal contexts .
it is woven into the research project, with significant consequences on the
outcomes.
., As social life becomes more saturated with Internet-based media for
communication, researchers will be able to creatively design projects that
utilize these media to observe culture, interact with participants, or collect
artifacts.
" Each new technology bears a double edge for qualitative researchers and
users; as it highlights or enables certain aspects and qualities of interaction,
it hides or constrains others.
In lnterpreting Qualitative Data, Silverman (2001) stresses the importance
of adhering to sensible and rigorous methods for making sense of data even
as we acknowledge that social phenomena are locally and socially constructed
through the activities of participants. Similarly, it is clear that although the
Internet can fundamentally shift some of our research practices by extend­
ing our reach, easing data collection, or providing new grounds for social
interaction, application of these methods must remain grounded in the
fundamentals of rigorous and systematic qualitative research methods.

N OTES

1. Much debate persists regarding the influence of the Internet on language use and
meaning. The vast majority of researchers agree that the structure and content of
CMC is distinctive. Language norms and rules are in constant flux and
transformation; time and space take on different meaning within interactions,
influenced by both technical and normative elements. It is unclear whether this
distinctiveness is meaningful at the level of meaning or discursive/relational
outcome. Early accounts suggested that the absence of non-verbal cues in CMC
would lead to less meaningful, surface interactions among users (Sproull and
Kiesler, 1991). Later researchers such as Witmer and Katzman (1998) find that users
make necessary changes in their discourse to accommodate technical limitations,
replacing non-verbals with emoticons. Gaiser (1997) goes further to contend that
there is very little difference between data collected in face-to-face and online
interactions. More recently, Thurlow (2003) argues that shortcuts used in SMS R EFE R E N C ES
(telephone instant messaging) do not significantly influence the meaning of the
message, although to an outsider witnessing the interaction, the discourse may seem Association of Internet Res
earchers (2002) Ethical deCl"�lon
from httPd
almost unreadable. Baym et al. (2002) contend that it is not so much the technology Retrieved December 1 , 2002 making and Internet research
5
. / /www.aOIr.o .
that influences interpersonal relationships as it is the interaction itself. Baym, N. (2000) Tune In Log rg/reports/ethi cs.p df
' On . oaps, Fan om, and Onlme Com
2. See also Sproull and Kiesler (1991 ) and Chen et al. (in press) for general perspectives. CA: Sage. munity. Thousand Oaks
3. See also the special issue on persistent conversation in the Journal of Computer­ B aym, N., Zhang, Y.B ., '
and Lin M (2002) The Int.er
�et 111
fer�nce of th e ASsocIatIOn of Internet
Mediated Communication, 4(4), 1999. presented at the annual con ' college social life. Paper

' 0ctober.
4. Heritage in this volume provides an excellent overview of conversation analysis, Maastricht, The Netherland Re searchers,
Bened1'l(t, M. (ed.) (1991) s
which seeks to examine and illustrate how context is accomplished in and through Cyberspace: First steps. Cam
bridge, MA : MIT Press.
1 20

121
ANNETTE N. MARKHAM TEXTS
INTER NET COMM UNICA TION AS A TOOL
FOR QUAL ITATI VE RESEARCH

Bromseth, J. (2002) Public Places . . . Private Activities? In A Morrison (ed.�, Researching Mann, C. (2002) Generating data online: Ethical
ICTs in Context (pp. 44-72). Oslo: Intermedia Report 3/ �002. Ump�b �orlag. Keynote address delivered at Making Com
concerns and challenges for the C21 researcher
,
. mon Ground: A Nordic conference on
Retrieved December 1, 2002 from http:/ /www.intermedla.U1o.no/pubhkasJoner / Internet research ethics, Trondheim, Norw
rapport_3/ ay, June 1 .
Mann, C. and Stewart, F. (2000) Internet comm
Issues W�en Teachmg
. . . and Learmng , unication and qualitative research: A handbook
Carvajal, D. (2002) The Artisan' s Tools. Cnhcal for researching online. London: Sage.
CAQDAS. Forum: Qualitative Social Research, 3 (2). Retneved Feb�uary 2, 2003 from Markham, A (1998 ) Life online: Researching real
http: / /www.qualitative-research.net/fqs-texte/2-02/2-02carvaJal-e.htm experience in virtual space, Walnut Creek,
CA: AltaMira Press.
Chen, S.1.S., Hall, J.G., and Johns, M.D. (eds.) (2003) Onlme Social Research: Methods, Markham, A (2000) Losing, Gaining, and Refra
Issues, and Ethics. New York: Peter Lang. ming control: Lessons from students ofonlin
courses. Paper presented at the second e
. . . international conference Learning2000
Cherny, 1. and Weise, E.R. (eds.) (1996) Wired_women: Gender and new realitIes m Roanoke, VA, October. '
cyberspace. Seattle, WA: Seal Press .
. onlme.. Markham, A (forthcoming) The Internet as
Danet, B. (2001) Cyberpl@y: Commumcatmg Oxford: Berg. research context. In C. Seale, J. Gubrium
. . tnckster
. . . G. Gobo and D. Silverman (eds), Qualitative
McLu�an, !VI' (1962 ) The Gutenberg galaxy: The
research practice. London: Sage .
'

Dibbell, J. (1993) A rape in cyberspace or how an evil clown, a Halhan spmt,

two wizards, and a cast of dozens turned a database into a society. The VIllage VOIce, making of typographic man . Toronto:
Umversity of Toronto Press,
December 21: 36-42. Miller, D. and Slater, D. (2000) The Internet: ,
Featherstone, M. and Burrows, R. (eds.) (1995) Cyberspace/cyberbodies/cyberpunlc: cultures An Ethn ographic Approach . Oxford:
Berg.
of technological embodiment. London: Sage. Nov�k, M. (1991 ) Liquid architectures in Cybe
.
Frankel, M.s . and Siang, S. (1999) Ethical and Legal Aspects of Huma� Subjects . �esearch First steps (pp. 225-54). Cambridge, MA:
rspace. In M. Benedikt (ed.), Cyberspace:
MIT Press. Quote retrieved April S, 2003
on the Internet. http://www.aaas.org/spp/dspp/sfrl/projects/mtres/mam.htm from http :/ /ww w,centrifuge. org/ marc os/
The Future of Feminist Internet Studies (2002) Panel discussion at the annual conference Persist�nt Conve�sation. Jour nal of Computer-
of the Association of Internet Researchers, Maastricht, The Netherlands, 15 ?ctober. Retneved Apn1 29, 2003, from http ://www.
Mediated Communication, 4 (4), June , 1 999,
Gaiser, T. (1997) Conducting on-line focus groups. Social Science Computer ReVIew, (15): ascusc.org/jcm c/voI4/is sue4 /
Postman, N. (1985 ) Amu sing ourselves to death
135-44. : Public discourse in the age of show business,
New York: Viking Press.
Gajjala, R. (2002) An Interrupted Postcolonial/Feminist Cyberethno�raphy; comprlClty ' Rose, E. (1960 ) The English record of a natu
and Resistance in the "Cyberfield," Fenllmst Media Studies, 2 (2). 177--93. ral SOciology. American Sociological Review
XXV (April): 193-208. '
Gergen, K. (1991) The Saturated Self. New York: Basic Books. Ryen, A (2002) Paper presented at Making
Giddens, A (1991) Modernity and self-Identity: self and soclCty m . the Late Modern Age.
on Internet research ethics, Trondheim, Norw
Common Ground: A Nordic conference
Cambridge: Polity Press. ay, June 1 .
. . LmgUlstlc, . . social and cross-
. . Senft T. (2003) Home page heroines: Gender,
Herring, S.c. (ed.) (1996) Computer-Mediated C?mrr:umcatwn: celebrity and auto -performance o n the Worl
WIde Web. Unpublished working doctoral d
cultural perspectives. Amsterdam: John BenJamms. dissertation, New York University.
. Sharf, B. (1999) Beyond netiquette: The ethic
Herring, S.c. (1999) Interactional coherence in CMe. Journal of Computer-Medlated s of doing naturalistic research on the
Internet. In S. Jones (ed.) , Doing Internet Resea
Communication, 4 (4). Retrieved April 1, 2003 from http:/ /www.ascusc.org/Jcmc/
. .
ShIelds, R. (ed.) (1996) Cult ures of Internet:
rch. London: Sage.
vol4/ issue4/herring.html virtual spaces, real histories, livin g bodie
London: Sage. s,
Hine, C. (2000) Virtual ethnography. London: Sage. Silverman, D. (ed.) (1997) Qualitative resea
Hudson, J.M. and Bruckman, A (2002) IRC Francais: The creation of an Internet based Sage.
rch: Theory, methods and practice. Lond
on:
SLA community. Computer Assisted Language Learnmg, 15 (2): 109:-34. Silverman D. (200 1) Interpreting qualitativ
. : e data: Methods for analysing talk, text
lTU (International Telecommunications Union) (2002) Internet IndIcators. Retneved mteractw n. London: Sage. and
April 1, 2003 from http://www.itu.int/lTU-D /ict/ statistics/ at_glance/InternetOl Soja, E.W. (1989) Postmodern Geographies: The
Reassertion of Space in Critical Social Theor
. pdf London: Verso, .y.
Johnson, C. (2003) Social Interaction and Mea�ing �onstructwn .
among �ommumty .
vvebsites.
TAT . Sondheim, A (ed.) (1996 ) Being Online, net subje
. ctivity. New York: Lusitania.
Unpublished Master of Arts thesis, UnIverslty �f IllmOls at ChIcago. Sproull, L. and Kiesler, S. (199 1) Connectio
. . . ns: New ways of working in the networked
Jones, S.G. (1995) Understanding commumty m the mformatlon age. In S.G. Jones (ed.), enVIronm ent. Cambridge, MA: MIT Pres
Cybersociety: Computer-mediated commumcatwn and commumty (pp. 10-35). Thousand s.
Stone, AR. (199 6) The war of desire and
,
Oaks, CA: Sage. technology at the close of the mechanica
. . Cambridge, MA : MlT Press. l age.
Jones, S.G. (ed.) (1997) Virtual Culture: Identity and commumcatwn m cybersoczety. London..
,

ThUrlow, C. (2003) Research talk pres

Sage. ented at University of Illinois at Chicago,
" , , February.
Jones, S.G. (ed.) (1999) Doing Internet research: Cnttcal lssues and methods for exammmg Turkle, S. (1995 ) Life on the screen: ident
'

the Net. Thousand Oaks, CA: Sage. ity in the age of the Internet. New York
and Schuster. : Simon
.
Watson R, (1997) Ethnomethodology
,
Kendall, 1. (2002) Hanging Out in the Virtual Pub: Masculinities and RelatIOnshIps Onlme,
'

Berkeley, CA: University of California Press. � and textual analysis. In D. Silverman (ed.)
QualttatlVe research: Theory, meth od and ,
practice (pp. 80-9 8). London: Sage .

1 22
1 23
 TEXTS
ANNETTE N. MAR KHAM w

t: GraphIC ' acc ents

and Kat zma n, S.L. (19 98) Smile when you say tha Sud weeks,
1
Witmer, D.F. . computer-med'lated communication. In F.
as gen der ma r <er s In . play: Vir tua l Gro ups on the
Rafael! (ed s. ) , Network and Net
M L McLaughlin, and S.
In�e;net (pp . 3-11 ). Menlo
ss.
Park, CA: AAAl / MIT Pre

"inside" the

Finding realities in interview,s

Jody Miller and Barry Glassner

In his Interpreting Qualitative Data, Silverman (2001) highlights the dilemmas

facing interview researchers concerning what to make of their data. On the
one hand, positivists have as a goal the creation of the "pure" interview -
enacted in a sterilized context, in such a way that it comes as close as possible
to providing a "mirror reflection" of the reality that exists in the social world.
This position has been thoroughly critiqued over the years in terms of both
its feasibility and its desirability. On the other hand, emotionalists suggest
that unstructured, open-ended interviewing can and does elicit "authentic
accounts of subjective experience." While, as Silverman points out, this
approach is "seductive," a significant problem lies in the question of whether
these "authentic accounts" are actually, instead, the repetition of familiar
cultural tales. Finally, radical social constructionists suggest that no knowledge
about a reality that is "out there" in the social world can be obtained from the
interview, because the interview is obviously and exclusively an interaction
between the interviewer and interview subject in which both participants
create and construct narrative versions of the social world. The problem with
looking at these narratives as representative of some "truth" in the world,
according to these scholars, is that they are context specific, invented, if you
will, to fit the demands of the interactive context of the interview, and
representative of nothing more or less.
For those of us who hope to learn about the social world, and, in particular,
hope to contribute knowledge that can be beneficial in expanding under­
standing and useful for fostering social change, the proposition that our
interviews are meaningless beyond the context in which they occur is a
daunting one. This is not to say that we accept the positivist view of the
possibility of untouched data available through standardized interviewing,

1 24
 THE "INSIDE"AND THE "OUTSIDE"
GLA SSN ER ' INTERVIEWS
JOD Y M ILLE R AND BAR RY
m subjective depth" and "deep mutual understanding" can be achieved (and,
tak e a rom ant "
1cl � e d VIe' w of sea mless authenticity emerging front with these, the achievement of knowledge of social worlds),
nor tha, t we It IS, to suggest tha.t we are not wil ling to, discou
narratlVe accounts ' Instead" rni ng ab ou t th e s ocial wo rld bey ond the mtervl' ew Those of us who aim to understand and document others' understandings
entirely the po ssib ilit y of lea choose qualitative interviewing because it provides us with a means for
in our �nalyses of intervie-: dantlt�i a p osition that is outside of this objectivist- exploring the points of view of our research subjects, while granting these
In thIS cha pter, we try to Ide Y , ous1y the goals and critiques ,of points of view the culturally honored status of reality, As Charmaz explains:
, poleS , We will ar ue that information about SOCIa1
, ' S t con I
fn uu m yet ta k es sen
constructlvi
bo th of ItS
'n-de th in�erviewing, The position we
researchers , at are We start with the experiencing person and try to share his or her subjective view,
wo rld s is achIevable thro�g� I , e!b authors such as Harding (19 87) and Our task is objective in the sense that we try to describe it with depth and detail. In
� ;
attempting to pu forw��d Is ��;� anti�dUalistic options for methodologi�al
doing so, we try to represent the person's view fairly and to portray it as consistent
Latour (1993), w 0 POSI , exp d'Ia stud'Ies and science studies - options WhICh
with his or her meanings, (1995: 54)

and th eon " zmg pr actices m me rses such

,
recognize that bot h emulat10 ' n and reJ' ect'10n of dominant discou rses are
Silverman and others accurately suggest that this portrayal of what we
,
vism m ISS somethmg " cnt'lca11y I'mPortant. Domin ant dis cou
fissures
do is in some ways romanticized, We will address below some of the problems
as positi , lete wit h
total'lzmg on1, y. for those wh'th'
, o VIew them as, such,' they are rephly satisfying and that make this the case, But the proposition that romanticizing negates, in itself,
w h'lCh people engage in hig the objectivity Charmaz defines, or the subjectivities with which we work,
and uncolonIzed spaces WI m e ma km, g, does not follow,
even resistant practices of knowt led wh
�
Ile We have no trouble acknowledging, for instance, that interviewees
We concur with San der s tha
sometimes respond to interviewers through the use of familiar narrative
[wl e would do we l� to heed
:he cautions offered by postmo dern ethnographers
about the bases of
constructs, rather than by providing meaningful insights into their subjective
� c� between being ske ptic al view, Indeed, as Denzin notes:
;
a con sId era ble dif fere
, , , [tlh ere is with these claims are
exan.un�n the rounds upon
tru th clai ms wh ile �are
fo �d e,d (a conventIOnal mte
ful ly
:
ractI011lS ente prise) and denyin
97 )
g that truth as a The subject is more than can be contained in a text, and a text is only a reproduction
ntation - exists at all. (199 5: 93, of what the subject has told us, What the subject tells us is itself something that has
uh�l Itanan and 1I'berating orie been shaped by prior cultural understandings, Most important, language, which
is our window into the subject's world (and our world), plays tricks, It displaces
the very thing it is supposed to represent, so that what is always given is a trace of
WO RL DS
NA RR AT IVE S AN D other things, not the thing - lived experience - itself, (1991 : 68)
, ervle rs in the interactionist tradition, inte rview
tes , for mt ' we In addition to this displacing, the language of interviewing (like all other
As Sil ver ma n no
no : , t arrat , v� s, but social wo rld s, For researchers in
Jus telling) fractures the stories being told, This occurs inevitably within a story­
uct �
subjec
this tra
ts con
dit ion
str
, "th e pn ma r�
r:
ISS U � IS � generate data which give an authenticts
, (S'1 verman, 200 1 ' 87) , While interactionis teller's narrative, which must be partial because it cannot be infinite in length,
insight int o peo ple 's exp e� len ces ' and all the more partial if it is not to be unbearably boring, In the qualita­
do not sug ges t tha t the re IS tive interview process, the research commits further fractures as well, The
coding, categorization, and typologizing of stories result in telling only parts
" or ab s o lute . yl do recogn, ize'
world out-t here" , " [the of stories, rather than presenting them in their "wholeness" (Charmaz, 1995:
"a sin gu lar obJ ech ve fication is essenhal lf
e obJ'ecti
.
"obJech" fled worId . eed ' they contend tllat som
s," Ind 60), Numerous levels of representation occur from the moment of "primary
ts, thus, not as an abso1ute
,
hu�an conduct IS t,O b cc m : � � ft �
lished Obj ecti vity exis
' d ion(to hich humans react
but as an accomplished experience" to the reading of researchers' textual presentation of findings,
or mherently meanmg u co including the level of attending to the experience, telling it to the researcher,
nce, Dawson and Prus, 199 1 5: 13)
aspect of human lived expene transcribing and analyzing what is told, and the reading,
, lection of the social world thatet Qualitative interviewers recognize these fissures from the ideal text (i.e "
Re sea rch cannot provIde the mirror ref attribu ' interviewees' subjective view as experienced by the interviewees themselves).
, tIV
OSI , stn' ve for, but it may provl' de access to the meanings people
, IStS a sym bohc Interviewers note, for example, that "[tlhe story is being told to particular
P , is itse lf
' l worId s, Wh1' le the interview dge of the SOCla, 1
Ia
to then �xpen' �nces and SOC, sco unt he ossibility that knowle y in the context
people; it might have taken a different form if someone else were the listener"
interactIon, thIS does not d: n can 0btam�
' ed, ' In fact it is onl (Riessman, 1993: 11), The issue of how interviewees respond to us based on
world beyond the intera�t1O s whIch . e recog�Ize nd build on their interactive who we are in their lives, as well as the social categories to which we belong,
contro an� re duce them) , that "inte
of non-positivis tic int erv iew ', r- such as age, gender, class, and race is a practical concern as well as an
components (rather than trymg to
127
1 26
 'o
T H E " INSIDE"AND T H E "OUTSI E "
RY GLA SSN ER INTERVIEWS
JOD Y MIL LER AND BAR
0

epistemological or theoretical one . The issue may be exacerbated,. for example,

when we study groups with whom
as a result of social distances, or
not understand our questions,lack of
ticularly
world itself. Ways of thinking and t 11 .
also used in these. (Glassner and Lo g � ���, d1��;' e3��5� daily experiences and are
we do not share members hip Par
interviewees may not tru d us iny their
st us, the may
they may purposely mis lea
mbership in their primary groups , LI F E OUTSI D E THE I NTERVIEW

responses . Likewise, given a ut theme phenomenon under stu dy to ask the

we may not know enough abo Interactionist research starts from a bel'Ief that people create and maintain
right questions. sents meaningful worlds As interact.�" omst researc� with adolescents illustrates,
Studying adolescents, as we s.hav e done in our own research, pre this belief can be �ccepted "WItltout assummg the eX1S. t ence , 0f a .smgle, .
unique concerns along these linethoOn
the one hand, the me ani ng sys tem s of
st encompassing obdurate realit " (Cl �
19 : 62). To assume th�lt realities
.
beyond the interview context �ann �:�;:p' ) � . mto �nd explored IS to grant
of adults, and adult res ear che rs mu
adolescents are different from theysehav g of ado lesc ent
e an understan din
exercise caution in assuming en there." On the other hand, adolescents are narrative omnipotence. The roots of th�se r:a Ihes are ,more fundamental and
cultures because they have "be oming increasingly oriented to adult worlds, pervasive" (Dawson and Prus 199 5.. 121, see also Daw�on and Prus, 1993�'
Schmitt' 1993) than such ew can account for. P:- vivid illustration of thi
. '

in a transitional period of life, bece and Sandstrom, 1988: 60). As a consequence,

though with "rough edges" (Finortance as a means of differentiating oneself, is to be
. found in Ch arm:z�� wor k on the chromcally· . . ill' inwher h0, she notes,
"age begins to decrease in impal differentiation, such as gender and class [and expenence sickness regardless of wI1ether th. ey partICIpate .interviews
and other dimensions of cultur e and Sandstrom, 198 8: 66). These dimensions (1995: 50). We note that the ad0lescents m our st�d'le� .�xpe�lence their
0 1 entity whether
race], become more crucial" (Fin establishing research relationships, rapport, age-, gender-, and ethnic-based identities and . fluid
. 1 Y
are thus of critical importance inh the information obtained and the interaction or not we interview them and WI'th'm our mterv1ews . . with them . and the
and trust, and in evaluating bot erviews. I�anguage sh apes meanings but also . per� mts mtersubjectivity
that occurs, within in-depth intage as the determinant or predictor of his or ability of willful persons to creat
. e an d mamtam meaningfu1 worId s ( Dawson
To treat a young person's and Prus, 1993 .' 166) . Recogmzmg tllIS' bwe can�ot accept the proposition that
neglect another key point about age
,

her experiences or ways of talking is to interviews do not yield informal' IOn a out socI�l worlds. Rather, "we take it
ordering as well: that two persons can commum. cate the1r. percephons to one another. Knowing .
full well that there are both .struc t ures .and polluta . n ts m ' any d'1SCUSSlOn
The idea of an ending of childho
od is predicated upon a normative
system wherein
we choose to stud what 's :�
. ����� � r
s�u:: � (Gl;�.sner and Loughlin:
chil dho od itself is taken for gra
nted. But childho od may also be "en ded " by
nstituting the 1987: 33). While c:rtainly �'tl�:�� I . . a1993 rea - 1ve person. between
narratives of personal or societal
"deviance" or by new stories reco the two covers of a text," as Denzm r ' ( m S chm1tt .' 130) puts It, we can
modelling of childhood itself. (Ro
gers and Rogers, 199 2: 153 ) describe truthfully delimited seg men ts 0f �ea.I-;1ve .
persons' lives. Indeed,
in so delimitin , we rnaY get closer to peo ple s lIved experience. As Charmaz
(1995) notes, rfany people d0 not want themselves' reveaIed. m. theIr. totality.
adolescents talk about in open-e nded
In our experience, mu ch of wh at see m wayward, delinquen t, pre mature,
interviews is precisely how their acts Recognizing this and respondin accordin?1� �� y �esult m dee� er, fuller
youthful behavior. Their discou y are. Itardis
rse tow
or otherwise not befitting proper is much about where and who the �onceptualizations of those aspect� of our subJec s 1ves we are most mterested
and with us (and for themselves)s and m understanding.
about trying out social location identities: �
Much the same deserves to be said abo�. th� ll' � terac :lOmst . . researcher
concerning the place and fullness of his or h � �l�hm the mterv1ew . context.
On the one hand, scholarship should pres:�:e , m It �he presence, concerns,
ents in social
lescents' reports as situated elem
Our approach is to treat the ado and to another
and the experience of the [researche 1 knov:er �nd dIscoverer" (Smith, 1987:
ways of making sense to one self
worlds. On the one hand they are much of
92; see also Harding, 1987) so tha� t�e . subjectIvIty
to reco gni ze that
(cf. Mills, 194 0). One cannot read
the transcripts and fail
ld con sider . that exists in all social
to understand topics that neither wou rviewees
what goes on is two persons trying inte research will be a visible part of the ��0�;� , and t��s available to the reader
:
for examination. As Harding (1987.' 9
in such special circumstances. The
in quite this manner or detail except , and often when the researcher appears
typically seem to enjoy the chance
to "think aloud" about such matters
is dire cted at a maj or project us not as an invisible , anony mo us VOIce. 0f 'authority , b u t as a rea l , . tonc. to
h1s al
ch of that thinking . "
they say this to the interviewer. Mu and wha t type .
of their present lives - figuring
out what type of person they are
ous pos sibi lities
md1v1dual with concrete, spec1'f'1C d eSlres' and mterest s" the research process
they want to be. The interview offe
rs an opportunity to try out vari can be scrutinized.
ing questions, and wit h refe ren ce to how it fits
Yet, on the other hand, these dictates do not nece� slta . :e, as some excessively
on this older student who is ask at oth er aud ienc es. On the othe r C

with one' s self-image or might wor

k out if directed
into the soci al revealing authors have taken them to mean, engagmg m confessionals either
and world come from and build
hand, these ways of viewing self
129
1 28
-

JODY MILLER AND BARRY GLASSNER ¢ INTERVIEWS

THE " I NSID E "AND T H E "OUT
6
SI E "

with one's interviewees or with one's readers, or boring them with excessive interviewees In our exp ' . .
details about oneself. It is precisely the "concrete, specific desires and interests" what will be�ome of th:���:��� er :
vle wees , principal concerns focus upon
that merit airing, not everything that might be aired. such as the protection of confidentia.litTh ose concerns extend beyond rnatters
In our experience, interviewees will tell us, if given the chance, which of they have to say matters. They want to��n�;��' wee� want to know t.hat what
I t
our interests and formulations make sense and nonsense to them. Glassner and A researcher who interviewed AIDS pa t'lent at wIll become of theIr words.
Loughlin (1987: 36) describe instances in their study in which the interviewer s 0bserved:
brought up a topic that was seen by the subject as irrelevant or misinterpre­ Ma�y of I�y �esp ndents expliCitly
tation, and they offered correction. Moreover, as Charmaz points out, "creating . � refer to their interviews as "legaCIes. .
partIclp atmg In Hns project despite the ain
it mi � t
h
" They are
t. C. use them because they
these observations at all assumes that we share enough experience with our
subjects and our readers to define things similarly" (1995: 64).
I will use their stories to help others.
T K .
of giving meaning to their lives and �s they s ou d �r me WIth the responsibility
believe
thelr d eath s. (Weltz, 1987 : 21)
Of paramount importance regarding how (and how much) we present
ourselves is the influence this presentation has on interviewees' ability and ���� f,ft�n' �he ��s7::rt �r b0th int:rVie�er and interviewee is less
willingness to tell various sorts of stories. Richardson notes, "People organize a n l p rtal t. ln m terVlews with ad QI escent . monu-
their personal biographies and understand them through the stories they create found, for example, that to be1 ta1(en . ous 1y and . s we have
to explain and justify their life experiences" (1990: 23; see also Lempert, 1994; sen
someone whose societal role is that of "teache . , re g r
. �g d e d as � teacher by
Mishler, 1986; Riessman, 1993). Highlighting two types of stories of particular characteristic of the interview situation. r " IS a def mm and hIg hly valued
relevance, Richardson first describes the "cultural story": "Participation in
a culture includes participation in the narratives of that culture, a general
understanding of the stock of meanings and their relationships to each other" AN I LLU ST RATIO N
(Richardson, 1990: 24). These narratives represent the basis on which indi­
viduals create cultural stories, or stories about social phenomena that are We have suggested that narratives which . rvl.
typically "told from the point of view of the ruling interests and the normative are situated in social worlds,' they come out emerge m ' mte ew contexts
order" (Richardson, 1990: 25). An interviewer who presents him- or herself interview itself W :r��:;� �nl� for of wo rlds tha t exis t out side of the
either as too deeply committed to those interests and that order, or as clearly /
also for our abilit a � the exi sten ce of the se wo rlds, but
our scholarshi T������:t;:o�e �f ��e
. rs 0 ca � tur e elem ent s of the se wo rlds in
outside of them, restricts which cultural stories interviewees may tell and how
these will be told. ing that accest�e mteractionist strategies for achiev­
se rc e ort one of us has recently com
Cultural stories are based in part on stereotypes. Richardson dubs an (Miller, 2001). pleted
alternative to these the "collective story." Collective stories take the point of The study involved in-depth . .
(ages 12 to 20) who claim affili:ti0!e:i��d d te::ew.s ��t� young wO.n:en
view of the interview subjects, and "give voice to those who are silenced
or marginalized in the cultural [story]" (Richardson, 1990: 25). They challenge By and large these an y:u�� g gs m elf commumtIes.
popular stereotypes by "resist[ing] the cultural narratives about groups of ;
male and fe�ale me�b�:s� ��� 1. �- ept
e ' tegr�ed, n:ixed- I?end er gangs with both
people and tell[ing] alternative stories" (Richardson, 1990: 25). A strength pletion of a surve interview adm ' h mtervIews followed the com-
InIstered by the same researcher' Wh 'l
1 e
of qualitative interviewing is precisely its capacity to access self-reflexivity th . !,
. ncle survey mtervIew gathered information b U . .
{.
1

among interview subjects, leading to the greater likelihood of the telling of � �ding the individual, her school, friends, fa�il� ��g���::���e o tOP ICS,
collective stories: :��o�;::��:������is��y' �e�ual histo:y, an? victimization, in 'ad;dli��:e��
t
Respondents may reveal feelings, beliefs, and private doubts that contradict with the roles and activ�tie;�f ye m-deptoh mt�rvIew was concerned exclusively
or conflict with "what everyone thinks," including sentiments that break the they describe as emerging frOI����i�g :en l� ���th gangs, and the meanings
Com d t h : g a i on.
dominant feeling rules . . . . In other cases, interviewers will discover the anxiety, ' ter��. :w:;� Mille� was t n to fifteen years
ambivalence, and uncertainty that lie behind respondents' conformity. (Kleinman
et al., 1994: 43)
;;
typicall :rce��e� :
often of a different rac!(Milfer is ���el��rvIe:ve�S), �f th� same. gender, but
older (although
African American) and class ba k!§oun' e maJon�y 0 the mtervIewees were
Here again, to be a candidate for "good listener," the interviewer does best middle, working class and P Oo�) . orn d (�Pter :lddle versus middle, lower
to present him- or herself as someone who is neither firmly entrenched in the should be members of th < e sc 0 a�s ave argued that researchers

�!:��:F; �
mainstream nor too far at any particular margin. Ultimately, though, it is not U I r t s
where the interviewer locates him- or herself that is of greatest relevance to ::
��:� ; u�����;;:; ;d : : ;:
,� ; � t��; ' �f� : �e�c::. ��r � �����
e egl Imate knowledge claims, researche
:
rs
1 30
131
 T H E " INSIDE"AN D THE "OUTSIDE"
RY GLA SSN ER INTERVIEWS
JOD Y M I LLE R AND MR

(19 90: 232).

�
stereotyping because I know, I know a few gang-bangers '
d their material in some fashion"95:
who go to school, get
should "have lived or experienceent straight A's, hit the books but still when theY o the street, you know, they take
36) point .
good care of theirs. They takin' care ,0f. the· 1rs sc 1001 and they takin' care of theirs
s specifically, Tay lor et al. (19
Likewise, with regard to adolesc ose a form of politic al res ista nce -t hat is, ' 111
on the street and J don't think that's rrght to stereotype people.
out that lI[a ldolescents may cho at the y kno w and fee l- to peo ple the y see as
to spe ak abo ut wh
choose not
ing or alig nin g wit h unr esp ons ive institutions and authorities." We have suggested that knowled g e of soClal " worlds eme�ges from the
represent
ing the se obs tac les can onl y be achieved by building rapport, achievement of intersubjective de th and mutu al understandmg. For , these
Overcom however' there m�st b e a level of trust between . the mter-
of the
ensuring and reassuring confide ntiality, and establishing trust ·- facets to. be present, '
research we describe in greatesug r detail below. s V Iewer and .
mterviewee . Socl
' a1 d'
IS t ances th at '
mclud e 1
d'ff � rences m relative
t the existence of social difference
.

Ho we ver , we wo uld also ges t tha

intervi ew s power can result in suspicion and lack of trus'�i �oth .of whIch the researcher
between the interviewer and inte rviewees does not mean that the ew must actively seek to overcome · RapP ort b dIng IS. a key
n abo ut soc ial wo rlds. In fac t, the inte rvi s can be
Establishing trust and famI'l'Ian. 'ty' sh owmg genume. mterest,
.m
. to this assuring
process.
oid of info rm atio
are dev
plis hed in wa ys tha t put the se soc ial differences to use in providing confidentiality and not bel'.ng JUd gmental are some. I�port ant elements
accom
ties for ind ivid ual s to arti cul ate their feelings about their life a f bmldmg. . '
rapport (Glassner and LoughI'�n, 1: 987.. �5) . .MIller found that the
opport uni ent ial benefit of social distances in
nce s. As not ed abo ve, one pot last of these was particularly important w en mtervIewmg adolescent female
experie
ch of this nat ure is tha t the inte rvi ew ee can recognize him- or herself asl gang members. These youn g women were member s' 0f a group frequently
resear
exp ert on a top iC of inte res t to som eone typically in a more powerfu s of
.
stIgmatized by the social groups' to who�c11 M'll � er h erself belongs, a reality
an
vis -ii-v is the soc ial stru ctu re (in this case, particularly in term known to both the interviewees a nd th e In t ervlewer
pos itio
, rac
n
e, and edu cat ion ). To find one self placed in this position can beabouth bot
M'll1 er' s research design proved usef�l m' allevIatmg . ' . tensions that could
age ause one can reflect on and speof ak . . .
empowering and illuminating bec mb ers groups result from this schI'sm . She b egan WIt h the admInIstratIOn of a survey
life in wa ys not ofte n ava ilab le. Wh en individ ual s are me interview, which began with reI a t'Ive 1 y .
mnocuous que s fIOns (d emographics,
one
t
's
hav e bee n ster eot ype d and dev alu ed by the larger culture, and wh), ose the
. . arrangements attitudes toward sCh00 1) , and slowly mad e the transition
hvmg
tha
es hav e bee n ign ore d (as in the case of female gang members from these to more s�nsitive questIOns ' about gang mv . �1vemen.t, del'mquency,
per spe
mis
ctiv
e of this app roa ch is all the mo re apparent. In fact, those interviearchs ew .
and VIctimization. In addition comPleting the survey mterview first allowed
pro her African American male res " WI'th eac h young w n:an, so t11at when the
conducted by Miller, rather thandifbyferent types2 of detail. her to establish a relationshIp
.m-depth interview was completed" there was a pre-exIstmg �
assistant, often elicited more and s from respondents often were ben eficial, level of familiarity. .
Miller found that her differenceTay lor and her colleagues suggest that one
In particular
. . ' in the course of th e survey . w 1len . a y
. oung woman descnbed
e tru st had bee n esta blis hed . ed partrclpating in serious delin qu ency -. m:ludmg, m some instances, brutal
onc
ant age of soc ial dis tan ce is tha t it "ma y elicit explanations that are assum g her acts of violence Miller had the opportumty to res.pond in a non-judgmental
adv h insider status (1995: 36) . Rea din
manner. This provided a layer of u.n� erstandmg that
_

to be known by someone witass ces . dfacilitated frank

II

transcripts against her research undista nt's, Miller found numerous instan ista nt, discussions when these issues w ere reVlSI t e d d uring th e In- ept h . .
mtervIews.
in which girls assumed sharedan and erstandings with the research assmunity �Iller also worked to build trust thro�g� ef�orts to protect youths'
.
had grown up in the same com 1 entIa l Ity IS ach'Ieved as' much by
because he was African Americre time to explain things to her. The res earch confIdentiality.
. . . assurances The assurance
. of conf'd
as them, while they took moitively knew what girls meant when they used ImplICIt as by explicit gu n tees. Th. ere were o�ten ?pportunities
assistant likewise felt he intu of events, and thus often did not ask for greater for Miller to convey her concern . ab �r:t protectmg the subjects. . .privacy. For
particular words or made note distances facilitated respondents' recognition example, when interviewin
. enile detention center), she remainedareas
g m open (such as the vlsI.tmg room at
explication. In this way, social ir social worlds. a Ju : aw r :�f he s�rroundmgs to ensure
of themselves as experts on the ls were cognizant of the controlling images a�aI�st eavesdropping, temporarily stopped�:e m erVIew when others came
In addition, many of the gir aspects of their lives. Consequently, in some wlthm clear hearing range and;:oved t� more secl� ded areas when neces­
(Collins, 1990) used to describe to Miller in ways that directly addressed and sary. Likewise, when int�rvie es �.ccI entallh dIsclosed names, Miller
�
instances young women talked s about adolescents, inner-city youths, and immediately offered to erase thes� S IpS r�m t e t� pe recording and did
challenged common stereotypewoman explained: so in the interviewee's presence b: ore �.ontmumg . wIth t�e interview. Most
.
gangs. For instance, one young importantly, she was proactive i ��ac I;g to the ImmedIate environment,
hardcore and and did not wait for signs of disco ort ro� respondents. In fact, she often
Some people stereotype, they just
stereotype gang members to be
that that could remained more concerned than theY were. lh. mately, she found that these
dy. They don't stereotype people actions helped to engender trust.
to always be shootin' at somebo stra igh t A's . That' 5
could go to school and get
be a gang member but still they
1 33
1 32
JODY MILLER AND BARRY GLASSNER INTERVIEWS
THE "INS IDE "AN D THE
d
"OU TSI E"

Cultural stories This young Wo ma n wa s frustrate . bY TV

Rapport involves more, however, than provisions of confidentiality, non­ perceptions of her non-gang pee d' n 0 t Ju�t depIc. tio
.
ns, but by the
rs . She continued :
judgmental responses, and other offerings from the interviewer . It involves
the interviewee feeling comfortable and competent enough in the interaction Girls at school . . . was like,
sexed in." . . . Likc, "oh, you
"well ou �O\�:
bably got se ed . n. You pro
� bably got
to "talk back" (Blumer, 1969: 22) - to label particular topics irrelevant, point your business . " They was like,
a ho
� Y o U get put I�;,: I was like,
"yo p r0bably got sexed m. "none of
out misinterpretations, and offer corrections. When respondents talk back they didn't . " . I was like, "no, .r reaIIy
provide insights into the narratives they use to describe the meanings of their
social worlds and into their experience of the worlds of which they are a Instead, describing her initiation, she
part. One way Miller's interviewees talked back - both to her and to the audi­ explained:
ences for her works about them - was in their resistance to her efforts to get I got y respect off
� the bat becaus ' , SIX
. [pu
took SIX from four o-uys I too . e' instead of takm
. d from four g ys . I got my res I
them to articulate how gender inequality shaped their experiences in the gang. nches] from the girls
off the bat. So the girls, 'they had
b k SlX h 1t S to th e hea
In doing so, they situated their personal narratives into larger cultural stories nothin' t0 say about me bem . � pect
the guys. a punk, neither did
about girls in gangs, both by vigilantly challenging these stories, and by
embracing them in their discussions of "other" girls . This type ?f ":alking back" was a
Scholars have noted "the impact of 'already established cultural standards' story about gIr direct challenge to one sor ' t of cultural
on individuals" (Schmitt, 1993: 126). Research participants gain at least part . ls In gangs that they are sexual
_
ly "1 oos e" nd functio . n as the
sex objects of male gang members In . on to de �
of their understanding, and convey their understanding to the interviewer, through her denial of being " sexe ' . add it � bu nkmg the reotype
by drawing on the narratives that come out of the social worlds around them . avenue of respect as being willin d In, " thIS young w oman d escste . ed her
nb
For adolescent girls, and gang girls in particular, a primary cultural standard g and abl e to WI . thstan .
peers. d a beatmg from male
they must contend with is the sexual double standard, in which they receive However, girls also drew from and " .
status from peers for their association with and attractiveness to males, but analysis and critique of "other" . 1 . ret1 lcated thIS cultural story in their
are denigrated for sexual activities deemed promiscuous. Scholars have noted One way they did so was b la���i m��d ing ?the� female gang members.
that a characteristic feature of adolescence is the "shift from the relatively themselves deemed sexuall 100 � a demgratIn� young wo me n they
�
asexual gender systems of childhood to the overtly sexualized gender systems sions of girls who had been sexed i�' fh' s was mo st eVIdent in their discus­
of adolescence and adulthood" (Thorne, 1993: 135), which has different really no ho pe for them . I mean It' ,On. ; you�� woman commente d, "there's
meanings and consequences for girls and boys. It is in this context that girls yourself." Another concurred ., "l'f yous Just tnflmg. ' tha t" s Just d ow ngrading
must negotiate - not just their sexual reputation - but their gender identity . . . They do n't care about theY bodY have sex With them, then you a wreck.
as well. off of them ." Describing gan g mem ' t��y Jus ' t let everybody get something
A long-standing cultural stereotype of girls' roles in gangs is that they are
peripheral or auxiliary members whose primary function is as a sexual outlet
.
sexed m, one girl explained "Th
·' �
b
y
ers
k no
mt
w
era
t h at
ctlO. ns w'I .
the y wa
I I
th glf S W 10 had been
s getting' looked at
as ho's . We J'ust look at 'em . Sorne
for male members (see Campbell (1984) for a discussion) . For the young ,
1ook, y all some little ho's, ' or 'why tIme s we tell ' em t 00, we be l'Ike, , ooh y'all
women in Miller's study, this cultural stereotype was further exacerbated by y ' a 11 d 0 t1 l aP' . "
irls �h� were sexually abused by ma
Though extremely judgmental of g
the practice within some gangs of "sexing in" some female members -
requiring the girl to have sex with multiple male members for entry into the
gang members, girls did not neg ativ ely ��a ua e young �en's sexual behaviorsle.
blame female victims of mistreat fam� l� cultural stones that single out and
In this way, gang girls drew from .
group (see Miller, 2001). In tandem with larger stereotypes, this practice
challenged the integrity and sexual reputations of girls in gangs . Thus, in sible for their victimization in bot:e�l . o;en continue to be held respon­
Miller's interviews, girls went to great lengths to distance themselves from Women and legal res onse � pu1ar Iscourses about violence aga
inst
the practice and to dispute this cultural story when describing their place in
the gang. One young woman expressed her frustration that people outside of them.
� i
When they described !nd ev'af:;!!� S cUl e� fr?m these cultural trad itions
'
e exp Oltahon of young women aro
und
gangs often assumed that she was sexed in:
Collective stor
ies
They be showin' these little movies on TV, like, well, the females have to get sexed
A prop OSI' t'IOn 0f thI'
in and the males have to get jumped in and like that. You know, you've seen 'em chapter is that l. ntervl. ews have
l.11teractional contextsSwithi the capacity to be
stOod. One way in which thins whICh. soc�l. worlds come to be better under-
on TV. And they, they just figure, well, if you a girl gang member then you got sexed o .
in . And I really didn' t. I wasn' t even down for nothin' like that.
1's ach leve IS throu
gh interviewees' revisions
1 34
1 35
 T H E " INSIDE"ANO THE "OUTSIDE"
GLA SSN ER INTERVIEWS
JOD Y MIL LER AND BAR RY
"

, as we have J.��t an sug ested We havesethe proposed that anound ther Finding realities in interviews
of cultural stories
an be t old as we e �.hat p�ivileg society. Miller's inter-er
soc ial wor ld We have suggested that a strength of qualitative interviewing is the oppor­
sort of story c '
. UssIO. n and 1. ts, .stories .over t.he st-ones af the larger tunity it provides to collect and rigorously examine narrative accounts of social
dlSC " . you
s1b1lIty ng women did more than just challennot ge worlds. In Miller's illustration, both cultural and collective stories provide
views illustrateeotth1Sypepos m e p1 � g theu
. . .
. vem e nt (1
gang mvo1 a collective story' . e ., "1 was important insights into how young women understand their place within
prevailing ster e resspec t :) . rns �� ad , the y also presented their gangs and within the broader social world of adolescent peer culture.
sexed in," "1 hav ce .
I ' We have also suggested that it is possible to find realities within interviews
of the gang as a spa �
n �1IT:� . qualit
' ::e ;�� : quite resistant to herthaatte
��y,wereme pha mpts - collective stories are a significant aspect of this reality, but so are accounts
The yo�ng w?me der me sizi ng inst ead t eve ry- of events and activities that are ultimately contradictory to these stories.
to elicit dISCUSSIOwanss "ofa11gen e ." One you ng woma n exp 1 'ned
aI .
' "the y give Despite their collective story of equality, without exception the young women
one in the gangof us resp�cthet thsam e wa :he� . the males " Anotheroff,was visibly Miller interviewed provided evidence to the contrary. For example, they
every last one Miller' s lIne af queYstIOns,��� rep eate dl� cut her or "tal ked described a distinct gender hierarchy within their gangs that included male
frustrated with se: leadership, a sexual double standard, the sexual exploitation of some young
back," in respon women, and most girls' exclusion from serious gang crime - specifically
and h �lf guy s?
. .
Miller. Yau sal'd before that
the gan g was abo ut half g'l rls
don 't thin k there
those acts that build status and reputation within the group. By juxtaposing
Can you tell me more a , b aut that ?. Like you said you girls' collective stories with these incongruous facets of girls' interviews, Miller
are any d ifferences in term
s of what -
(2001 ) built her theoretical discussion around the contradictory operation
Intervie wee: There isn' t!
of gender within gangs.
MI'/Ier' Ok ' can you tell me
mo re -
isn't , ther e isn't l i ke, ther e's noth
.
ing - b oy, For example, we noted that the practice of "sexing" girls into gangs was
Intervie
.
wee
. L'k
. l e w hat? . The re
, Chi nes e. one that challenged the integrity of all girls in mixed-gender gangs. As one
girl , whi te, blac k, Mex ican
non-gang girl told Miller, "I mean, they tell you that [they weren't sexed in],
ed, "if she conber
A nother girl notlIke sider herself as aclud gang m,;mber, t�tenreas�ye but you don't know how they really got in." Because of this, gang girls'
. g mem " ,and she con at.ed, you can vilification of girls who were sexed in was necessary in order to distance
gonna be treated a gan . . :
am t b ase d on like dudes, theisey, anot
th " Likew her themselves from the practice, and maintain an identity as a "true" member
judge. it o� I�he b.�� ��ethedudguless 'toIt the We just trea t us as someone who had not been sexed into the gang. The more vocally girls
explamed. WeweJ act so much like dudm. y can't do not hin g. The y resp ect
like that ' cause ugh, but we J"ust so mueschthe l 'ke
1 dudes that they just don't trip
spoke of these other girls as "ho's" and "wrecks," the more successful they
were, or hoped to be, in creating a rigid dichotomy between themselves
us as females tho and girls who were sexed in. Ironically, denigrating girls who were sexed in
off of it." suggest' part of girl s' discussions of theishor equ ality within or exploited, rather than holding the young men accountable, was the key to
As these commensotsmu ch by suggestings from that all wo men uld be treated maintaining the tenuous but vital belief that the male gang members viewed
the gang was notdifferentiat the. .mselve you ng wo me n who were not them as equals. Creating this dichotomy allowed girls to conclude that
equa11y, but by girl madeing explICIt: the boys in their gang treated and discussed other young women in sexually
in gangs. As one derogatory ways -young women who deserved it because they were "nasty."
:-
, anna break their nails and stuff like that. So,
On the other hand, they could believe that the boys considered them, as one
�\ : :
t
A lot of girls get scared. Don
d
ain 't no nee , for the r to y o
. n ng And the ones that 's in it, most
�:� :�;
th�y in, 'cau t
of
and young woman put it, "one of the niggas," or one of the guys.
the girls that s m act 1 (e oy .
T s se they like to figh
To challenge the practice, or the general tendency of some gang boys to
stuff. They know how to figh t and they use guns and stu ff.
mistreat girls, would be to risk being ostracized from the group precisely
ective story that situated because of its male-dominated structure. It would also mean having to
Thus, the y�ungthewogan men Miller spokebleto totoldthea coll
mal e mem bers in terms of confront rather than deny that girls were not "equals" in the eyes of their
ir
the . pla ce m g as � o �p ara ording to this collecti ve male gang peers. These young women lived in social worlds in which women
acf1V1t'1es, toughn ess an d W1 ll mg nes s to figh t. Acc . were devalued; their means of resisting their own devaluation was not to
' . hich theY receive status and esteem from b elllg
story, the gang. 11s. a:�;l��a 1�:tan d up for themselveers, tha exhibiting traits that challenge the premise of this treatment, but instead to define themselves as
n females. outside of its boundaries. Thus young women appreciated the normative space
���:�a����:� ��mmonry associate with males rath of "equality" available in gangs, even when it was not always a reality in
practice. Identifying with cultural stories about women, and rejecting such
137
1 36
 THE " INSIDE"AND THE "OUTS DE" I
RY GLA SSN ER INTERVIEWS
JOD Y MIL LER AND BAR
Charmaz, I�; (1 995) "Between positivism and postmodernism: Implications for
s for the ms elv es, allo we d gan g gir ls to con str uct and maintain the .
image methods, Studies In Symbolzc Interaction, 17: 43-72.

collective story that they were "one of the guys. "

.
Collms, P.H. (1990) Black Feminist Thought. Boston: Unwin H yman.
ethn ograp 1ly and postmodern
' approaching human rIved experience,"
Dawson, L.L. and Prus' R . C . (1993) "Int erac tlOnlSt "
.
d Iscourse: Affinities and disJ'unctures m
.
Stu d'les In Symbolic Interaction, 15: 147-77'
" :
. .
r
CO NC LU SIO N Dawson, L.L. and Prus, R . C . (1995) "Postmod ermsm a nd rmgUlshc ' reality versus
tha t, wh ile "op en- end ed" interviews candatbea use ­ . .
Silv erm an (20 01) arg ues fro m the nat ura lly occurring ntifiest tha
symbolic interactionism and obdurate rea lty, Studzes In Symbolic Interaction, 17:

ful, we nee d to jus tify dep art ing nti c" impulse whichrdsidein scare
105-24.
.

sur rou nd us and to be cau tiou s abo ut the "ro ma Denzin, N.K. (1991) "Representing lived expenences ' m ethnographic texts", Studies
h different wo
"experience" with "authenticity." We agree, but wit
in Symbolic Interaction, 12: 59-70.
.
Douglas, M. (1 986) Risk Acceptability According to the SOCIaI Sclcnces ' . New York: Russell
.
quotes. have tried to sug gest in this chaope pter some strategthaiest Sage Foundation.
Fine, G.A. and Sandstrom, K L (1988) l'L(nowlng Chlldren.. PartI"C)pant Observation with
On the one haniewd,swe
. .

wh ich int erv can be les s-th an- pro ble ma tica lly n-ended, and nce with
.
• .

by ntifying experiepro
Minors. Newbury Park, CA: Sage.
nee d not res ort to rom ant icis m, or to ide .
int erv iew ers experiences anduld pudut cein
Glassner, B. and Loughlin' J (1 987) Drugs In Adolescent Worlds: Burnouts to Straights.
cal l up on int erv iew ees ' New York: St. Martin's P;ess.
aut hen tici ty, in ord er to s. On the otherwehanqud,estweionwothe grounds .
authentic accoun ts of social worlddat a," because cultural, in sociological
Harding, S. (1987) Feminism and Methodology. Bloommgton: Indiana University Press.
" I1egmg '
.

scare quotes "na turally occurring en the natural gand

Kleinman, S. ' Stenross, B ., and M
.
. erviews:
. cM ah on' M ' (1994) "PnV
. .
freld work over

for any neat her distinction betas,we198 In ma kin suc h claims, it is notorthe case
mt Consequences for identity and practice," Symbolzc Interaction, 17 (1):

dat a as els ew e (d. Do ugl 6).

rely local eventsted. Instearesd,s
exp �: �
37-50.
Lat � ' (1993) We Have Never Been Modern. Cambridge, MA: Harvard University
that we are "noern t too sure whethaserSilv interviews are pu111 sugges
underlying extinstalthe realities," imperm an (2001: ) has y the m as one or the
s
.
dualistic erative to cla ssif Lempert, L.B. (1994) "A narrative analysis of abuse.' Connectmg the personal, the
rhetorical, and the structural " 10u al of Co mporary Ethnography, 22 (4): 41 1-41 .
;�
we argue aga �
Miller, J. (2001) One of the Guys .' Girls , angs an . ender. New York: Oxford University
other. s. Some comeatfromamtter other people, som e
All we sociolomgisour ts have are stosrie . Wh s is to und ers tan d Press.
.
Mills, C.W. (1940) "Situated actions and vocabuIanes of motives," Amencan Sociological
from us, someere fro interacare tion with others , wh ich sor t of stories they areut,
. .

how and whcan put the stories pro duced

them to honest and intelligent use in theorizing abo
ReView, 5: 904-13.
. .
Mishler, E.G. (1986) Research IntervieWing.. Context and Narrative. Cambridge, MA:
and how we
�:;�
Harvard University Press.
social life. .
RIC son, L. (1990) Writing Strategies: Reaching Diverse Audiences. Newbury Park, CA:
.
Riessman, c.K. (1993) Narrative Analysis. Newbur Park C . a
N OT ES �:� � i � � ��
RO � .s. and Rogers, W.S. (1992) Stories ofChil lOod. T�ro � : iversity of Toronto
St. Louis, Missouri. s
e study in Columbus, Ohio and .
1. The project wa s a comparativ were conducted by Miller; many of the interviews
l � :�
Sanders, c.R. (1995) "Stranger than fiction ' Insi h d pItfalls in post-modern
All of the Columbus interviews et�nography," Studies in Symbolic Interacti�n I ' 8 �
in St. Louis were conducted by
a research assistant. See Miller
(20 01) for an in-depth
SchmItt, R.L. (1993) "Cornerville as obdurate reajIty.
.
. . Retoolmg .
'
the research act through
cess.
des cription of the research pro postmodernism " Studies in SymbOIIC' In teractlOn, 15: 121-45.
Silverman, D. (200 1 ) Inter reting Quaiztatlve Data: Methods for Analysing Talk, Text and
and felt quite safe
arc h assistant as a paternal figure,
�
2. Girls respon ded to the rese ting , as this was not
with him (and withou t pro mp
gth
��
ing at len Interaction. London: sa e.
a dIca� cntique o� sociology," in
disc losing and talk ion. One girl even
ut their experiences of victimizat . . .
part of the research protocol) abo
:�
t. Smith, D.E. (1987) "Women's perspective
r tried to fix him up with her aun
asked him to adopt her; anothe S. Harding (ed.), Feminism and Methodol y . oommgton: IndIana University
Press. pp. 84-96.
.
��
Taylor, J.M., Gilligan, c., and Sullivan' A: M ' (1995) Between VOice and Silence: Women
and Girls, Race and Relationship Ca b d ge� MA: Harvard University Press.
RE FE RE NC ES
Thorne, B. (1993) Gender Piay: Gi�ls a oys In School. New Brunswick, NJ: Rutgers
Perspective and Method. Ber
keley, CA:
' as 1 egacy," Hastings Center Report, 17: 21-3.
Interactionism: University Press.
Blumer, H. (196 9) Symbolic Weitz, R. (1987) "The intervlew
.

University of California Pre ss.

the Gang. New York: Basil Bla
ckwell.
Gir ls in
Campbell, A. (1984) The
139

138

int erv iew
Jam es A. Holstein and Jab er
F. Gubrium
Thinking about the interview as beinwg as "active" is somewhat uncon�ent�onar­l.
a neutral means �f extractmg mfo
Typically, we approach the interviens. Res pondents provide a��wer s. !he
mation. Interviewers ask questio duit between .the t�o pa;tl.CIpants. [he
interview process is merely the concess keeps the mtervlewer s l�volven:ente
standard vision of the interview pro posed to be neutral, inconspIcuOUS, httl
to a minimum. The interviewer is sup speak.
more than a "fly on the wall," so toely at the i terv i w pro cess to see . how
In this chapter, we will look clos � �
ctic e - for bot h the mtervlewer and the respondent.
be any other .way, �o ma�ter how har
acti ve inte rvie win g is in pra
Indeed, we will argue that it cannot r presence m the mte rVIe W
interviewers might try to diminish thei ive Interview." Thi s title doe s not
Thus, we title the chapter "The Act ,
erentiating it from y � sa , the tand
out a distinctive type of interview, diff directive life stor y mte r �lew
survey interview or the minimallythat interactional, inte r� retl ve
we use the title to highlight the fact rviews are active inte rvIe ws.
a hallmark of all interviews. All inte
EN ESS
FO CU SI N G ON ACTIV
Today, interviewing is more popular than ever as a mea�s of generati�g,
information. In our postmodern "intervi ew society" (see Gubnull.l and Ho�stem
2002a, 2003; Silverman, 200 1), the ma ss media, human serVIce proVIders,
and researchers increasingly generate etele data by interviewing. The numberper
sho s, an� new spa
television newS programs, daytimresul�svisi �n tal�( �
wit h the of mte rvle� s IS gr0:-vmg by le.aps
articles that provide us Ical forms of mfor� atlO� co�lectlOn,
and bou nds . Regarding more method t of all social science mvestIgatIonstedl
it has been estimated that 90 percen (Briggs, 198 6). Inte rvie w ing is un� ou�
interviews in one way or another for . .
ducting sy�te�atlc s?c:�l mqUlry, .
the most widely used technique polcon ts, psychIatnst�, ��m�CIans, admm­
sociologists, psychologists, anthro treaogis interviews as theIr wmdows on the
istrators, politicians, and pollsters2002a;t Hy man et al., 1975):
world" (Gubrium and Holstein, vides a way of gen ra�m g emp �
. l data
. nca
In simple terms, inte rvie win g pro �
about theIr lIves. In thIS respect,
about the social world by asking people to talk
(
THE ACTIVE INTERVIEW
interviews a�e special forms of conversation. While these conversations may
:rary f:om hIghly .structure�, standardized, quantitatively oriented survey
mtervlews, to semIformal guIded conversations, to free-flowing informational
exchanges, all interviews are interactional. The narratives that are produced
may be as truncated as forced-choice survey answers or as elaborate as
oral life histories, but they are all a product of the talk between interview
participants.
While most researchers acknowledge the interactional character of the
inter:riew, t�e t�chnical literature on interviewing stresses the need t,o keep
that mteract�on m check. Guides to interviewing - especially those oriented
to standar�lzed s.urve�s - are primarily concerned with maximizing the
flow of valId, relIable mformation while minimizing distortions of what
the respondent knows (Gorden, 1987). The interview conversation is framed
as a p otential source of bias, error, misunderstanding, or misdirection; it is a
persls�ent set of probl�ms to be minimized. The corrective is simple: if the
mtervlewer asks q�eshons pr�perly and the interview situation is propitious,
the respondent wIll automatIcally convey the desired information. In this
conventional view, the interview conversation is a pipeline for transporting
Just knowledge.
A recently heightened sensitivity to representational matters (see Gubrium
d and Hol�tei�, 1997) - characteristic of poststructuralist, postmodernist,
exc h � .
nge const�uctlOmst, and ethnomethodological inquiry has raised a number of
sl�gle questIOns about the very possibility of collecting information in the manner
ard ized the conventional approach presupposes. Attention has increasingly focused
. I �. �
s tea on the activene�s o� inter.views. In :raried ways, these alternate perspectives
actI VIty IS ho�d that meamng IS SOCIally constltuted; all knowledge is created from the
actIons undertaken to obtain it (see, for example, Cicourel, 1964, 1974; Garfinkel,
19�7). Treating interviewing as a social encounter in which knowledge is
actlvely constructed suggests the possibility that the interview is not so much
a neutral conduit or source of distortion, but rather a site of, and occasion
for, producing reportable knowledge.
Sociolinguist Charles Briggs (1986) argues that the social circumstances
of interviews are more than obstacles to respondents' articulation of their
particular truths. Briggs notes that, like all other speech events, interviews
of fundamentally, not incidentally, shape the form and content of what is said.
Aar?n Cicourel (1974) goes farther, maintaining that interviews impose
pa�hc�lar w�ys of �nderstanding reality upon subjects' responses. The general
pomt �s that mtervlewers are deeply and unavoidably implicated in creating
use meanm� s that ostensibly reside within respondents (also see M anning,
y 1967; MIshler, 1986, 1991; Silverman, 2001). Both parties to the interview
as are n�ces.sarily an� unavoidably active. Meaning is not merely elicited by apt
questIOnmg, x:or �lmply transported through respondent replies; it is actively
and commumcatIvely assembled in the interview encounter. Respondents
�re not so much repositories of knowledge - treasuries of information await­
m� excava�ion as they are constructors of knowledge in association
.
WIth mtervlewers. Interviews are collaborative accomplishments, involving
141
JAMES A. HOLSTEIN AND JABER F. GUBRIUM INTERVIEWS
THE ACT IVE I NTE RV
I EW
participants in meaning-making work in the process (Alasuutari, 1995;
KEY
Holstein and Staples, 1992; Holstein and Gubrium, 1995).
Technical attempts to strip interviews of their interactional constituents
will be futile. Instead of refining the long list of methodological constraints In today's interview SOCle. ty, m
. form ' .
way of interviews. atIon IS increasingly acquired by
under which "standardized" interviews should be conducted, we suggest
that researchers embrace the view of the interview as unavoidably active and A common view of the int
begin to acknowledge, and capitalize upon, interviewers' and respondents' transporting knowledge. erview is that of a one-way pIp . elI. ne
for
constitutive contributions to the production of interview data. This means Recently we have come to recogm.
. g con
' ze the interview as a meanmg . -
consciously and conscientiously attending to both the interview process and makm versation a site an d occ�slO . n for
more like a two..wa y inform aflOn ma .
kmg meaning It is
the products that interviews generate in ways that are more sensitive to the
pipeline. a street than a one-way .da ta
social construction of knowledge.
. use 1't IS' a two-way conversat '
Beca
-f lI�.t
Conceiving of the interview as active means attending more to the ways . .
in which knowledge is assembled than is usually the case in traditional unavoidably interactional and con ion ' ervlew' mg I.S always
is active. struc Ive - m a word, the mt erview
approaches. In other words, understanding how the meaning-making process ,

unfolds in the interview is as critical as apprehending what is substantively

asked and conveyed. The haws of interviewing refer to the interactional,
narrative procedures of knowledge production, not merely to interview
techniques. The whats pertain to the issues guiding the interview, the content OF I NT ER VIE WI N
G
of questions, and the substantive information communicated by the respon­
.
dent. A dual interest in the haws and whats of meaning production goes hand Taking the activity of all int iew
in hand, expanding an appreciation of the constitutive activeness of the sections discuss how the interv ing as 0ur po mt of departure, the following
" erv '
l ew cu It'
Iva tes me ani . by
interview process. thIS· VISIOn of the interview in relatio . s ng. W . e begm locating
This appreciation derives from ethnomethodologically informed, social examining alternate images of the s:b? �; � t no con ceJ?tIOns of interviewing,
constructionist sensibilities (d. Berger and Luckmann, 1967; Blumer, 1969; (also see Gubrium and Holstein, 200 e behmd the mtervlew . pa
In order to find out about anothe 28 .
rticipants
Garfinkel, 1967; Heritage, 1984; Pollner, 1987). The process of meaning
ences We typically believe that w r p er� o�' s feelmgs, thoughts, or experi-
.
production is considered to be as important for social research as is the mean­
and �he other's reality will be ours e V to ask the right uestions
ing that is produced. In significant ways, this approach also resonates with
� S� �: ; er �e t, the consummate Jou �
methodological critiques and formulations offered by feminist scholars (see
DeVault, 1990; Harding, 1987; Reinharz, 1992; Smith, 1987). Ethnomethodology,
and mterviewer, says he sim 1 tur
/�
to talk. Writing of the intervfe�s n� �l IS �pe reco�der and asks people
.
he i or IS bo ok trtled Working,
rnalist
constructionism, poststructuralism, postmodernism, and some versions (1972: xxv) notes: Terkel
of feminism point to issues relating to subjectivity, complexity, perspective,
and meaning construction. Still, as valuable and insightful as they are, these There were questions, of cou
rse. But they were casual m.
"linguistically attuned" approaches can emphasize the haws of the interview would ask while haVing a d
rm' k WI'th someone·, the k'md
nature . . . the kind you
process at the expense of the whats of lived experience. We want to strike short, it was a conversation In he would ask you. . . . In
a balance between these haws and whats as a way of reappropriating the were open. " time the slUlce' gates of dammed up hur
ts and dreams
significance of substance and content to studies of interviewing. While
the emphasis on process has sharpened concern with, and debate over, the As straightforward and preten . Te , .
epistemological status of interview data, it is important to not lose track of vIe. wing permeates the soCIaun I sCIe
'
tious as It' IS,
. rke1 s Image of int
what is being asked about in interviews and, in turn, what is being conveyed "prospecting" for the true fac .
nce ' .w�. ng IS. generally likeneder-
s ' Intervle
by respondents. A narrow focus on the haws of meaning construction tends Of course, there is a hig hly sopts and fe�lings resldmg wI. thm . the to
hIS' t'Icated techn 0Iog y t1lat tells resres po nd en t.
to displace the significant whats meanings - that serve as the relevant grounds how to ask question
s the rts �f qUestIO s not to ask
. ear che rs
ask them, and ways t� avoIso
-

for asking and answering questions. � , the order in Which to

or otherWise bias the data (' d saymg the th mgs that mI.ght spOl' 1, contaminate
The basic modet howeverH '
1 7 Fowler and Mangione, 199
adroitly. ;=�:: : i \
i��1 ar 0 th e one Terkel exploits 0) :
so

1 42
143
-

V
THE ACTIVE INTER IEW

JAM ES A. H O LSTEIN
AND JABE R F . G U BRIUM
" INTERVIEWS
"neutral" than the interviewer who does not' The successful Impiementation .
soc ial sci en tifi c pro spe cto r iew as a searchy
casts the intervg wh of neutral practices elicits truths h ld m' the vessel of answers behind the
The im ag e of the iew er int en is alread
t on detectin lies inatext
and dis cov ery mi ssi on , wi th the int erv
nd en ts. Th e challenge Highly ref racting respondent. Validity results if unbia:ed P�oc,edures are su �ces�fully applied.
there ins ide va ria bly coo pe rat ive res po
con taminating it.the process. ine d In the vessel-of-answers a roach th I�f: of the subject ls, passi:e; this
ati on as dir ect ly as po ssi ble ,
inform hnologies streamline, systematize, and sanitize wi tho ut Th is subject is not engaged in th!�rodu�tio� . owledge. If the mtervlewing
interview tecvarying degrees of standardization (see Gubrium and Hoizelstdeinby, process goes "by the book" and IS' , non-d'lreciIonal and unb'lased, respondents
can involve by and Maccoby, 1954), ranging from interviews organ asure­ WI' 11 val'Idly speak out with whatev�r IS. presumed to reside inside themselves
,�on creeps
2002b; Maccospecifically worded questions and an orientation to meanized the unadulterated facts, of expenence : Cont am '
� m principle,inthefrom
nat
: teractlon; the
interview setting, its participants� and th lr m subject
_

structured, example, Singleton and Straits, 2002), to flexibly org jective is "pure, and, under ideal c� dj tion:' s, erves up authentic, reports, when
ment (see, gu for re ge ne ral qu est ion s aimed at uncov ering sub en, II

beckoned to do so,
int erv iew s ide d by mo na and Fre y, 2000; Johnson, 2002; W arr
me an ing s (e. g ., At kin son , 200 2; Fo nta What happens, however, if we enliven the �' mage of the subject behind the
respondent ?
if we view the sub'Ject he or she IS as active " From th'IS pe rspec­
2(02). dge contrasts what he calls "formative" with "mass" interveiewsors,t tlv, e, the subject behind the respondent not only holds facts and details
_

John Ma m according to whether the respondent "is given som y are of experience but in the very process of offering them up, constructively
categorizingtothe cho ose the top ics to be dis cu sse d and the way inn-dwhireich the
er­ add. s to, takes"away from. , and transfonTIS them mto ' artifacts of' the �ccasion,
of fre ed om
discussed" (19d 65: 165). Formative interviews ger include the noormal interve int
ctiv
iew s, Tlle �espondent can hardly "spoil" what he or she IS," m. effect, subjectively
ling (see Ro os,the194ma 5), inf
views favore iesin. Ro gerian counsesur int erv iew
s fall int vessconstraints and aims cat ego ry. creatmg,
and life historclassiMoficstatilaronge-censcaterles onvey This activated subject pieces ex eriences �gether, before, during, and after
By and large,iew process, with little att the communicidatito how interviews differ assuming the role of respondent. ' �s a mem er of society, he or she mediates
pa
of the interv for knowledge production.ention and alters the knowledge that ls conveyed t,o the interviewer; he or she is
0
"always already" an active maker f meamng. Be, �au�e the :esp�ndent's
as occasions answers are continually being assembled and modIfIed, mcludmg hIS or her
rticip an t sense
. of what lies within' tlle answers' truth value canno t be J'udged simply
m terms of whether they match wh t r es m ' a �essel of objective answers,
d th e interview pa
ch
l of the researin, From a more traditional standp�in:f th�, O:J�ChVIty
The subject behin
the typ e of int erv iew , the re is alw ay s a mo de lste , , or truth of interview
Regardless ofg beh erview roles (Gu
ind persons pla199ced5).inFointr exa brium and Hobeh 0
responses might be assessed in terms re la Ihty, the extent to which
subject lurkin
ein and Gubrisenumse, of epistemologmp tin g a sub
le, projec the respondentject ind
wers wh enever and whe rever 't
I '
IS carned, outques-
2002a; Holstent fer s a ica l age ncy up on a­
, tioning.
va1 Idity
yields the same ans . " YIelds the "correct" answers (Kirk
, the extent to wh'ICh mquny ' and
the respond onconour understanding of the relative validity of the inform ' " IS viewed as a dynamlc, '
which bearsreported. In conventional approaches, respondents are subdirjecectts
tion that is
and Miller' 1986) ' When tlle mtervlew
rna1(lng ' occasion however', d ' fferent cntena
1 " apply, The focus ismeaning- on how
ssive vessels ofof ans wers to whom interviewers meaning is constr�cted' the CIrcumstances of construction, and. the meaningful
basically conceis. veThdeyasareparep osi tor ies fac ts, ref lec tio ns, opinions, andervothiewer l'mkages that are assembled for the occaSIOn ' ' While interest . m' the content of
their questionerience. Occasionally, such as with especially sensitive int that it answers persists, it is primaril in how and, what the subJ,ect/ respondent, in
traces of exp th recalcitrant respondents, researchers acknowledge less, collaboration with an equally a�ti,ve m' terVIewer, produces and conveys about
topics or wi ult to obtain accurate experiential information. None theby the the subject/ respondent's expenence under the interpre t'Ive CIrcumstances
'
may be diffication is viewed, in principle, as held uncontaminated vide at hand (see Altheide and Johnson' 1994) ' 0ne cannot expect answers on one
the inform ssel of answers. The trick is to formulate questions and proween to replicate those on anotller because , they
subject's vehere conducive to open and undistorted communication bet occasion
,
CIrcumstances of production, Similarly, the V� h�Ity . ofemanswers
erge from d',Ifferent
denves not
an atmospewer and the respondent. from their correspondence to meanin�s �:l� wlt��n t�le respondent, but from
the intervi the methodological literature on interviewing deals with the their ability to convey situated expenen Ia realItIes m terms that are locally
Much of intricate matters. The vessel-of-answers view cautious inter­s comprehensible,
nuances of theesecare in how they ask questions,lest their manner of inquirysbiafor �he active image of the interview is best ut I'n perspech' ,:,e by contrasting
viewers to takthin the subject. This perspective offers myriad procedureiewer It, ,wIth
what lies wi adulterated facts and details, most of which rely upon interver who
obtaining un neutrality. For example, it is assumed that the interview more
more conventional images, " We will tof!the
dIffer considerably in thfeIr' onentatIOns
Us on two"claSSIC exemplars that
. expenentla1 truths held by
the passive subject , The Irst approach onents ' to the rational, factual value of
and questionions that acknowledge alternative sides of an issue is being
poses quest 145

1 44
JAMES A. HOLSTEIN AND JABER F . GUBRIUM INTERVIEWS
THE ACT IVE INTE RVIEW

what is communicated. It focuses on the substantive statements, explanations, Throughout, nverse and Schuman's b��1� provI�' es ,
and reasons with which the respondent articulates experience. We use Jean problematic the Co ghmp,ses of how
Converse and Howard Schuman's candid book Conversations at Random (1974) tions repeatedly ima
sho
ge of the r: assive subJ'ect
w us tha t m t erV '
Iew s
,
are
IS
con
m pra
ver
ctIc
sati
e, The nch illustra-
as an exemplary text. The second approach orients to the purportedly deeper on
are not only conveyed, but coo erativel bu'Ilt. up, ece� whe�e meamngs ' '
and more authentic value of the subject's feelings. It emphasizes sentiment and and recorded by interviewe; While rhe veteran �I � lv d, mt erp reted,
manage the pressuthe
.

emotion, the core of human experience. We use Jack Douglas's book Creative res of ' � ter VIe we
ing to sta d d' e conversabon for t�e purposesto r lea rns
Interviewing (1985) to illustrate this perspective. of inquiry, readers cannottry
hel
cal step" away from encounteripngbutthfeee '
m
�th:: t�:y are only a short eplstemologi-
' as an actI've,. meam. ng-makm, g
terVIew
Survey interviewing
occaSIOn,
While Converse and Schuman discuss the most standardized of interviewing
techniques, their book tries to portray the survey interview "as interviewers interviewing

see it," richly illustrating how interpretively engaging, and relatedly difficult Converse and uman's view of survel in er Iew
and exasperating, the survey respondent can be, The book describes the inter­ approach exempSch . . ontrasts wit
lified in Douglas'S b?Ok } ca/lVe �ntervlll:ewg, �mg, h the
esting and complex personalities and meanings that interviewers encounter we also see some marked sI'm' lan't Ies The word " crea t'lVe but, at bottom,
1
while interviewing, depicting them as "the pleasure of persons" and "con_ title refers prima to the interVIe' wer, not ' the responden m Douglas's
vle, wm. g derivesrily
1/ '

noisseurs of the particular." But the authors caution the reader that, even
though it will be evident throughout their discussion that the respondent can from the difficulties he enc?un tere� attt H"mp . •. IS ':"lew f ll1te
tmg to
0' r-
respondents' "deep exp
•

nce' " DougIas wntes that m hIS: manY empI" lcal probe
tUd'les, he repeatedly diserie
c .

be quite interpretively active, this does not work against obtaining objective S
information. This information, the reader eventually learns, is derived from cov
were for conducting researcherel·ntderVhow shallow the standard recommendatIOn �
the repository of knowledge that lies within the basically passive subject as those Converse and Schuman e p Iew · s, Canons of ratio na1 neutral'lty, suchs
behind the respondent. The authors do not believe that the respondent's his respondents' "emotional well:P��ge� f,�I'IInedrestopon capture wh,at Douglas calls
conduct implicates his or her subject in the construction of meaning, As Douglas calls for a methodology f deep d'ISC
0
. losure se to thIS shortcoming,
lively, uninhibited, entertaining, and difficult as the respondent might be M ost basically, as's difficulties relate s r; ' ,:"ch to hIS, Ima
at times, his or her passive subject ultimately holds the answers sought in passive subject as theDoy ugl
do to sho rtco ; ,
, ge of the
gs of stan�ar mtervlewm, g technique,
the research, Like the image of the subject behI'min d th
Converse and Schuman grant that survey interviewing involves experi­ imagin, es his subJ'ects to be repoSI�ton. ese sfurvanseyweresrspondent, Douglas also
0
encing the "pleasure of persons," but the authors hope that interviewers use are VIewed as but . in h' case, they
IS
their clearly designated roles and the rules of standardized interviewing who authenticawellyll-gcomuarded vessels of feel'mgs, HIS" model
s fr°m an emotIO. nal wels llsp of respondent
behest of an interviewermuwhnico katenow ring, at
a

to effectively gain access to the vessel of answers behind the respondent. Their cannot d raw o�ttheor
book is replete with anecdotal reminders of what interviewers must learn in convey what experience ultimately iss tallllatabomeutre Stwoanrds
order to keep that vessel of answers in view and the respondent on target and answers touch only the surface of e pen" en e, Dodard survey questIons
In part, it is a matter of controlling oneself as an interviewer so that one does intimately "ge , tting to know" the real sub�J ect beh�md theuglresp as aims deeper by
not interfere with what the passive subject is only too willing to put forth. CreatIve' mterviewing is a set of techniques for movm' g ondent,
The interviewer must shake off self-consciousness, suppress personal opinion, words and sen
interviewer mutenstces exchanged in the m' terVIew , process, Tobey ach
ond the mere
and avoid stereotyping the respondent. Learning the interviewer role is also
a matter of controlling the interview situation to facilitate the candid expre­ establish a cl'Imate for mutual disclos�re, Thiev
should be an occasion tha e '
e this,, the
mte rvlew
ssion of opinions and sentiments. Ideally, the interview should be conducted or her own feelings and deep t displahs th���' t��Ie, :v�r,s WIllIn
. gness to share
in private, This helps assure that respondents will speak directly from their that they can in turn share est t oug , ' . IS IS one to assure respondenhists
vessels of answers, and not be significantly affected by the presence of others. interviewer's deep dis�losure thbo��' �;�s:�nshmanateleg th��g�ts and feelings, The
The seasoned interviewer learns that the so-called pull of conversation, reciprocal revelations This D U 1as u gests, IS �hoItImIzes the respondent's
�
which might have an interpretive dynamic of its own fueled by the active the cultivated neutraiity of th� s�an;ar� survey rou,ghly suppressed by
subjects behind both interviewer and respondent, must be managed so that cardinal rule, he writes: mtervlew, As if to state a
the "push of inquiry" (p. 26) is kept in focus. Ideally, the cross-pressures of
conducting inquiry that will produce "good hard data" are managed by means Creative interviewing, as we sha
of "soft" conversation Cp, 22), strategies and tactics of interact
.
ll see thraughout, l11v
ion, la rge1y based on an underst
olves the use of many
anding of friendly
1 46
147
 THE ACTIVE INTERViEW
UM
AND JAB ER F . G U B R I
INTERVIEWS
JAM ES A. HOL STE IN
'\l

search
tua l disclosure and a creative
A close look at the traditional model of survey interviewing
, ' ' e, mu
Imize cooperatIV
fee l'111gS and intimacy, to opt
' , (p ' 25 em
tua l un d erstan dmg
. , the ongmal)
"

for mu
pha sis in
that, in practice, interview conversations hint that the respondent
interviewing, Onese99is toperfigur e more than a mere vessel of answers, Nevertheless, attempts
Douglas offers a set of �U1d, �hn' cre
es fo� creat, ivevie olv cen t standardization help sustain the vision of the passive subject.
that, as, h. e pu ts 1' t" "gemusthe m ative ttoerdeewiplyngdisinv
m
clo se req uir es mu c h In contrast, creative interviewing rests on a model of a deep
persplYatlOntha g res"pond,'ntA second admoni
" (p. 27) ' "gettmg mere agmg
tion for engyself" .
emotional subject, but this, too, is a vision of a relatively
o
e

more work npro obta1m� � �: :: �

a ul" is "reint searc her , know th subject or vessel of answers - a wellspring of feelings.
in "deep-de�puousbes sl;-an nto �he, the par t of the erviewer, who def usually
(p. 51) . Contmearcher,�IS neces a ySlS on creative interv iew er's ow n ense
is also the, res work against mu , y, lest"the and understanding. A thi
sar rd
mech. amsms a commI" tment totuamul dtuaISC11dis osure res sin g an a b'd'
I mg
sure· by exp iewer wh "has
rule IS to sh. oWelings. Referring tO � ne�Phyt, clo
I NTERVIEWIN G

e interv 0
interest in fe n derously revealmg l1fe studiesativ e cre
," Dougl as writes thatrmgs, the Years ago, Ithiel de Sola Pool (1957), a prominent critic of public opinion
polling, argued that the dynamic, communicative contingencies of the inter­
done some w�Iewer IS. ,, dn' ven by " friendly, cari ng, and ado ring fee view literally activate respondents' opinions. Every interview, he suggested,
creative intervthose an endean, ng, WI: de-eyed sense of wonderment at the is an "interpersonal drama with a developing plot" (p. 193), This metaphor
b t adds to veiled before her" (p 9) " conveys a far more active sense of interviewing than we see in traditional
l':ysteries un ' ngs tapped by crea't�lVe . m t e rvi ew ing are sai d to be emotional, models, It is an image of the interview as an occasion for constructing, not
The we11spn ast to the pref rred af o 1 'mage that permeates Converse merely discovering or conveying, information. As Pool indicated:
in distinct contr � ; , : ;: � e and wisdom are "partially
book: As, Dougc'�:o����f muotualedg
and Schumanof's cre l sea rch es for understanding,
atI ve mt era
the product mo, ns" ( 55 emphaSIS m on" gma1) ' While Douglas's imaginede
" The social milieu in which communication takes place [during interviews] modifies

of soul commusically emPotl' , On' a\�h' � b' t in the role of respondent, noniz not only what a person dares to say but even what he thinks he chooses to

subject is ba ly coop rates WI t�e I�t�;�i�wer to create mutually recogn "­

say, And these variations in expression cannot be viewed as mere deviations from

the less active s. In thI�s �ega d, the ��t�al"ty closure - the "crateatis ive
some underlying "true" opinion, for there is no neutral, non-social, uninfluenced

: - me la es,Iaddofs dis shapes wh said

situation to provide that baseline , (p , 192)
able meaning ve intervlewmg to , and
ati
thrust of cre ht. Conceiving of the interview as an interpersonal drama with a developing
in its own riguglas does not rec, ogmz'. e is that this admittedly active subjecter plot is part of a broader vision of reality as an ongoing, interpretive accom­
What Do c nstitute the ellsp , n f ex erience in rational or oth plishment. From this perspective, interview participants are practitioners of
could just as well � otl. Ona�one . �Y �:i�ing fa recognize this, the subject everyday life, constantly working to discern and designate the recognizable
ne ces
terms, not las's ressan ly em ;�ms : essentially passive, if creatively and orderly features of the experience in question, But meaning-making is not
ug po n � ent rem
behind Do nt f eX en�nc�, �o :1�r the respondent who "opens up"n merely artful (Garfinkel, 1967); meaning is not built "from scratch" on each
emotional, foug a dn �nk wIl?th tu s er ; , �ocated deeper inside, perhaps, thage interpretive occasion, Rather, interpretation orients to, and is conditioned by,
while havin hind the su�vey res�o; ert' the subject behind Douglas's ima of the substantive resources and contingencies of interaction,
the subject bendent remams a re a lVe y static, inert vessel, in this case, In other words, meaningful reality is constituted at the nexus of the haws
of the respo wers. and the whats of experience, by way of interpretive practice - the procedures
emotional ans and resources used to apprehend, organize, and represent reality (Holstein,
1993; Holstein and Gubrium, 1994, Gubrium and Holstein, 2000). Interviewing
is a form of interpretive practice involving respondents and interviewers
as they articulate their orientations and understandings in terms of what
, Garfinkel (1967) calls "practical reasoning," Reality is continually "under
All app roa che s to int erv iewing pr,e�umet.some age or model of the
nn .
construction," It is assembled using the interpretive resources at hand, in light
. t beh'md tl1e interview" partlClpan
subjec ind participants as of the contingencies of the moment. Meaning is not constantly formulated
Traditional approaches enVISlOn the subJ'ect beh anew, but reflects relatively enduring local contingencies and conditions of
possibility (Foucault, 1979), such as the research topics presented by inter­
passive. viewers, participants' biographical particulars, local ways of orienting to those
149
148
JAMES A. HOLSTEIN AND JABER F. G U BRIUM INTERVIEWS
THE ACT IVE INTE RV EW
i
topics, institutionalized ways of understanding and talking about things, and
other accountable features of "what everyone knows" about a topic (Gubrium,
1988, 1989; Holstein and Gubrium, 1994, 2000a, 2000b; Gubrium and Holstein,
2000). Those resources are astutely and adroitly crafted to the demands of the
occasion, so that meaning is neither predetermined nor absolutely unique.
Active participants

The image of the active interview transforms the view of the subject behind
participants. The respondent is transformed from a repository of opinions and
reason or a wellspring of emotions into a productive source of either form
of knowledge. The subject behind the interviewer is similarly activated. From
the time one identifies a research topic, to respondent selection, questioning
and answering, and, finally, to the interpretation of responses, interviewing
itself is a meaning-making project. The imagined subject behind the participant
emerges as part of the project; it is not presumed to exist before or inde­
pendently of the interview conversation. Within the interview itself, the
subject is fleshed out - rationally, emotionally, in combination, or otherwise the whats and hows to
-in relation to the give-and-take of interviewing and the interview's research Interviews, of course d n
work
.
purposes. The interview and its participants are constantly developing.
Two sets of communicative contingencies enter into the interview's they the only occasi�nsholWh �1n;��j!c�/1 a��e��:it;:;pr����s,�raemo ctic�. Nor are
meaning-making activity. As noted earlier, one kind involves the substantive reports are socially constructed Why th n, IS" mtervIe. wI. g a tIons, and
whats of the interview enterprise (see Gubrium and Holstein, 1997, 2000). The
ful venue for systematic sociai in �ir � � .n espe�ial1y use-
focus and emerging data of the research project provide interpretive resources s�tuation's sr: ecial ability to conce�ed i;; i��;e ��:;�����t�� t mter�iew
for developing both the subject and his or her responses. For example, a project � at a�dress ISSues relating to particu lar rese arch n ��. meamngs
con cern
mterviews expressly put the whats and ho f ' t . ' . sll:nple terms, s I n
might center on the quality of care and quality of life of nursing horne residents �n �I t�nda:dized interviewing, the passi:es :u�:�;�:�:�:i1;rachce to w?rk.
(see Gubrium, 1993). This might be part of a study related to the national s e
debate about the proper organization of horne and institutional care. If i� :��;�: !:� ;:���t���1::r::;i���:��ct��ei:t::���;;��ch:1e �
conception of the interview how ver, .nvests the ubject wit g f. �;ur�;:qt�
interviews are employed, participants draw out the substantiality of these r
topics, linking the topics to biographical particulars in the interview process, ac Ive�
producing a subject who responds to, or is affected by, the matters under repertoire of interpretive �eth �s an� an extens�Ive stock h a substantial
0
materials. of experiential
consideration. In one case, a nursing horne resident might speak animatedly
tap�� �: ;!��r t�:�C:i��t�!��:=�1;�::e�1e ima
during an interview about the quality of care in her facility, asserting that, "for o ge,of. the vess:l waiting to be
a woman, it ultimately gets down to feelings." This brings to mind Douglas's nd
are activated, stimulated, and cultivated�m t�nte ll� 1terviell1ter.pretIve capabilities
emotional subject, articulating a recognizable linkage between affect and interview is als? a commonly recogm. zed occasion for formw exchange. The
gender. Another resident might coolly and methodically list her facility's aticall y provokmg the respondent to formulate and talk aboallyutand sy.stem-
qualities of care, never once mentioning her feelings about them. Offering . .
�f�:; s �e : mot�ons in. particular ways, implicating the intervie;�;,r����:
o an e e
her own take on the matter, the respondent might state that "getting emo­ � . Ctl've InterVlewers d not merel
tional" over "these things clouds clear judgment, implicating a different 0 h.
kind of subject, more like the rational respondent idealized in Converse and
II

Schuman's text. Particular substantive resources -such as the common cultural

�����e;: S� q
;:���: a� ;�:���t�: :�
VIewers may sugges:�:;��teorie
t
£i:�::i:?r���E���:�
link between women and feelings, the traditional cultural opposition of r ntations to and linka b . '
clear thought and emotionality, or institutionalized ways of dealing with x erience,
:��t�:�� �:����;:���i: ;eS: ; nnectio hinting �t - even inv���g ���::;�e���i��:
emotional problems which are promoted by, say, psychoanalytic or twelve­ may explore incomplete1 r� ����t�� aspectsns,ofandexpout .
ene
looks. Interviewers
step therapeutic methods - are used to form the subject. respondents to develop rOpi·CS m. ways relevant to the nce, encouraging
ir own experience
1 50

151
JAMES A. HOLSTEIN AND JABER F. GUIlRIUM $ INTERVIEWS
THE ACTIVE INTER IEW V
(DeVault, 1990). In some interviews, the objective is not to dictate an inter­ !
k ds and my husb and. Peop le are
pretive frame, as a minimalist standardized survey approach would do, but living long er now and you 've g ot
I:
different gene ratio ns at hom e and,
to provide an environment conducive to the production of the range and How do you feel about it in your situa
I tell ya, it's a mixe d bles sin g
.
complexity of meanings that might occur to all interview participants. R: Oh " I don 't know Som
tion?
etim es I think I ' m bein g a bit selfis
Pool's dramaturgic metaphor is apt because it refers to both the whats I g npe about haVi: ng to keep an eye h beca use
on Moth er all the time . If you let
and the haws of the interview. As a drama, the interview conversation is dow n your guar d, she wan ders off
into the back yard or goes out the
scripted in that it has a topic or topics, distinguishable roles, and a format door � nd dow n the street. That 's no
fun whe n your hubb y wan ts your
for conversation. But it also has a developing plot, in which topics, roles, and atten tion too. N orm work s the seco
"
, d, ay a Iot. I man age to get i n a few hours of work , but he does
nd shift and he ' "," hom e d unng '
format are fashioned in the give-and-take of the interview. Viewed in this way, the
l i ke It. I have pretty mixe d feeli ngs n 't
the interview is a kind of limited "improvisational" performance. The produc­ I:
abo ut it.
What do you mea n?
tion is spontaneous, yet structured focused within the loose parameters �
R: Wel l, I d say that as a
provided by the interviewer, who is also an active participant.
daug hter , I feel pretty g u ilty abou
som etim es. It can get pretty bad , t how I feel

While the respondent, for one, actively constructs and assembles answers, gone , you know what I mea n? She'
:
l i ke wish ing that Mother were just
s been a won derf ul moth er an d
'
he or she does not simply "break out" talking. Neither elaborate narratives I lov her very muc h, but if you ask
me how I feel as a Wife and moth er
nor one-word replies emerge without provocation. The active interviewer's that s anot he� matter. I feel like she'
, s [the moth er], well, intru ding o �
role is to stimulate respondents' answers, working up responses in the process. our l ives and Just mak ,ing hell out of raiSi ng
a fam ily. Som etim es I put
Where standardized approaches to interviewing attempt to strip the inter­ myself In my husb and' s shoe s and
I just know how he feels. He does n't
view of all but the most neutral, impersonal stimuli (but see Holstein and say muc h, but I know that he m isses
cour se. [Pause] So how do you answ
my com pany , and I m iss his of
Gubrium, 1995 and Houtkoop-Steenstra, 2000, for a discussion of the inevit­ er that?

able failure of these attempts), an active sense of interviewing turns us to the The interviewe� then goes on to explain that the respondent should answ
. the
narrative positions, resources, orientations, and precedents that are brought 111 way she bel best represents her thoughts and feeling s. But as theer
into play in the process. The interviewer attempts to activate an appropriate exchange unfolds, It�evesbecomes evident that ''best'' mIs. represents the complexIty .
stock of knowledge (Schutz, 1967) and bring it to bear on the discussion at �f the respondent's thoughts and feel ings . In the follo win g extr act
hand in ways that fit the research agenda.
Consider, for example, how diverse aspects of a respondent'S knowledge, �e :esp
dIstm ct Iden �� struggles to sort her responses to accord with c�noti
.ondetitIe te g
ce how
oric all. y
s. At one point' she explains that she now knows how a WIfe
perspectives, roles, and orientations are activated and implicated in an couId and s hould feel because she gathered from the way her husband and
interview involving an adult daughter who is caring for her mother -a victim sons acted that "men don 't feel things in the same way. " Th' . IS suggests that
of senile dementia -at home. The daughter is employed part-time, and shares her own th oughtsand feelmgs ' are draw n from a fund of gen dere
the household with her employed husband and their two sons, one a as welldNote, too, .how at several points the interviewer collaboradtes knowled e
part-time college student and the other a full-time security guard. The extract respon ent to defme her Identity as a respondent. The interviewer inwith t�e
.
begins when the interviewer (I) asks the adult daughter (R) to describe her also p ts the wha and haws of the interview to work. At the very nd �
feelings about having to juggle so many needs and schedules. This relates pr.eced��ng extract,ts.the resp ond ent sug gest s that othe r resp ond ents
� �: ��
to a discussion of the so-called "sandwich generation," which is said to be �I��t serve to clanfy the wa� �he h�rself organized her responses, ' answers
caught between having to raise their own children and seeing to the needs of a mterper.sonal contextuahzmg mIght encourage even more interpindi reta
cating
frail elderly parents. Note how, after the interviewer asks the respondent of her expenence. She elaborates this position as she continues. tions
what she means by saying that she has mixed feelings, the respondent makes
explicit reference to various ways of thinking about the matter, as if to sug­ R: I try to put m�self in their [hus band
and sons] shoe s, try to look at it
gest that there is more than one plot to the story. The respondent displays from their pOin t of view , you know
, from a man 's way of thin kin g I
considerable narrative activeness; she not only references possible whats of ask myself how it feels to have a part
-time wife and mama I ask m '
care giving and family life, but, in the process, informs the interviewer of how ���
how I 'd feel. Belie ve me, I know he ' lf
I�. [hus band ] feels pretty rotten a
t
she could construct her answer. Men get that way; the� want wha
time , �ell, they , re qUie t, l ike noth
t they want and the rest of the
ing's the mat ter. I used to thin k
was gOin g crazy with all the stuff on I
I:
my mind and havi ng to thin k about
We were talking about, you said you were a member of the, what d id everythin g all at once and not bein
g able to fin ish with one thing and
you call it? get on to the othe r. Yo � know how
it gets - doin g one thing and feeli ng
R: They say that I ' m in the sandwich generation. You know, l i ke we're bad abo ut how you did som ethin
g else and wan ting to redo what
sandwiched between having to care for my mother . . . and my grown you did or what you said . The way
a woman does , I gues s. I think I 've

1 52
153
 ,
THE ACTIVE INTERVIEW
JAME S A. HOLS TEIN AND JABE
R F. GUBR IUM " INTER VIEWS

t know . It's pretty com plicated

learn ed that abou t mys elf. I don'
how do I really feel? Because all interviews involve the active construction of
:
think ing abou t it. [Pause] Let's see,
reality, the traditional model of the respondent as a vessel of
Pynprlpntl
.
I g, you men tione d bein g sand wich ed earlie r
Well , I was just won derin
and wha t a wom an feel s? . and the interviewer as a neutral interrogator loses its appeal.
I gues s I was n't all that sure what wom en l i ke me feel u �tll
boys felt. I figur ed pretty q � lck
R: Yeah
'
I figu red out how Norm and the
thing s out and that, well , I !ust
th�t men are pretty good at sorti ng
t feel thing s the sam e way . I .Just
coul dn't do it, ' caus e, well, men don'
wou ldn 't want to do that way anyw
ay. �
Wou ldn't feel r ght about It as
I N TERVIEW DATA

a wom an , you know what I mea n? So, like they say, live and let live,
Compared to more conventional perspectives on interviewing, the active
I:
I gues s. view might suggest that interviewing merely invites unacceptable forms
But as a dau ghter?
ask me how I feel havin g Moth er unde
r foot of bias into the information gathering process. After all, in this view, far
R: Yeah that too. So if you
� r not so far back that I was � nder more is going on in interviews than simply retrieving information from
all th time , I ' d say that I reme
foot a lot whe n I was a little girl
mbe
and M othe r neve r com plain ed, respondents' repositories of knowledge. "Contamination" would seem to
and she' d help Dad out i n the store
, too. So I gues s I coul d tell you lurk everywhere.
I'm healt hy and arou nd to take care of her and, hone stly, This criticism only holds if one takes a narrow view of interpretive practice.
Bias is a meaningful concept only if the subject is seen to possess a preformed,
th at I'm glad
don' t know. You' ve talke d to othe r
I ' d do it all over agai n if I had to. I
wom en about it . What do they say? purely informational commodity that the interview process might some­
�: � how taint. But if interview responses are seen as products of interpretive
�� just thinking that m �ybe
ell,
on't want to put you on the spot. I was practice, they are neither preformed, nor ever pure. Any interview situation
if f�� ew how others in my shoe s felt, I m ight be able to sort thing s
- no matter how formalized, restricted, or standardized relies upon
interaction between participants who are constantly engaged in interpretive
_.

out better than I did for ya.

comments about both the subject matter undthat er con­ practice. Because interviewing is unavoidably collaborative (Garfinkel, 1967;
The resPondent'sone or should. formulat e resp onse s show the Sacks et a1., 1974), it is virtually impossible to free any interaction from those
Sl'd erat10' n and' how boradoes . , '
actlV at es 1ve
d' rse
with the mterviewer. g quesflOns and �nswers . com- factors that could be construed as contaminants. Participants in an interview
resp, ondent, ln collas as antion angm are inevitably implicated in making meaning. They are involved in meaning
municative resource w as active,ralwepartcanof exch
integ a�d . apprec�a�e 1 .w 10 construction, not contamination.
Viewin the intervie icipates with the respondowlente�mge sluf
ackn pOSltlOn� In While naturally occurring talk and interaction may appear to be more
the int:rviewer part explore alternate perspectives and stocktmg :v�e ge. spontaneous or less "staged" than an interview, this is true only in the sense
the interview so as to the general parameters for responses, c?�kno s of
��s that such interaction is staged by persons other than an interviewer. Seemingly
The interviewer setsanswers that are germane to the researcher s mteremm stra
� t. e spontaneous conversations are not necessarily more "authentic" or bias-free.
well as provoking t is discussed is partly defined by the research tOpIC and They simply take place in what have been conventionally recognized as
per t·mence of whatantive horizons of the ongomg . m. terVl' ew exchange. non-interview settings. But these settings, too, play a definite role in the pro­
partly by the subs duction of experiential knowledge - just like interview situations. Still,
with the development of the interview society, and the related increasing
deprivatization of personal experience (see Gubrium and Holstein, 1995;
KEY t"'ViN I � Holstein and Gubrium, 2000a), the interview is becoming more and more
a developing pl?,t." commonplace, increasingly making it a naturally occurring occasion in its
The interview is an "inte rpersonal drama withrper sonal drama �re own right for articulating experience.
The active interviewinpart icipants in this "inte king - th,: ongomg Given this new, active sense of the interview, how does one make sense of
constantly engaged gniz the work of meaning-mayday expe lence.. interview data? Once we acknowledge that all interactional and discursive
production of the reconingfulable features of ever from �mter s data are products of interpretive practice, analysis may center as much on
The versions of mea experience that emergeand the whatview of the haws as on the whats of interviewing. For example, traditionally, interviews
are constituted in .the interplay of the haws s have been used in service to naturalistic research, focusing on what social
worlds are like (see Gubrium and Holstein, 1997: Chap. 2). They are typically
interpretive practice analyzed as more or less accurate descriptions of experience, as reports or
155
1 54
&

V
THE ACTIVE INTER IEW
R F. GUBR IUM INTERVIEWS
JAM ES A. HOLS TEIN AND JABE
N

rally, re-presentatio�zmg s! of reality. A:na.lysis,s atake s the .fo.rm

representations (lite ng and summan apstheulatdesescnp tlOn nd
. pro vidmg
of systematically grongupifram ework that enc pondenand exp lam s. asp ec�s .of The concept of the active interview casts interview "bias" in a new
a coherent organizi ondents portray. Resreport; the whats of expene�lvlty
ts' inte rpre tlve aC �ight. All parti�ipants in an interview are implicated in making mean­
the social world tothattheresp
substance of what they nce mg. They are mvolved in meaning construction, not contamination.
is subordinated over the . The leading question should not be whether or not interview
take precedence archer haw s.
reco gniz es the acti . behmd
. vIty ' the pro duc tlon procedures contaminate data, but how the interview generates useful
When the rese e data may also be analyzed to describe how interview information about the phenomenon of interest.
of interview data,asthos l as to show the dynami c interrelatedness oftsthearewhanotts Becau�e interview data are products of interpretive practice; data
talk is produced, intewel w conversations. fixe Respondents' comInstmen analysIs demands a new sensitivity to both the interview process and
and the haws of reporvie vered from a d repo sitory. ead, they are the unfolding substance of responses.
viewed as reality wayrtss deli that they construfocu ct aspects of exponeriethentiaasse l reality
considered for the the interviewer. The s is as muc h mbly
in collaboration with mbled.
process as on whaticalis lyasseorie forms of r:arratlve . an� dIsc ' ourse analY�ls,' .
Using sociolog rds of intented rpretive practlce aremea exammed to revealcirc reahty­ CON C L U D I NG POINTS
conversational reco es as well as the subjecti ve ning � that are u�­
constructing practic(see er, 2002; DeVa,-:lt,sma Gub nu � and Hol stem , �n a final no.te, ,:"e want to emphasize several leading points of this chapter.
stantially conveyed andBak 1990 ;
n, ; Silv
Gubrium, 1994; Riesare p�odu�ed in the mteractlOn er � an, 20� 1). Flr.st, the active mterview is not a particular type of interview, to be distin­
1994, 2000; Holstein
1993
respo�ses gUIshed from other forms of interviewing. We use the term active interview
The goal is to showwer and resprvieondwent,
how inte
wIth out losm g sIgh t of the mea mng. s to un�erscore. the perspective that all interviews are unavoidably active
between intervie circumstances that mediate the mea�ing-making pr?cess. me.anmg-m�kmg ventures. Even the standardized survey interview itself is
produced or the ctive is not merely to descnbe . the sItuated pro.ductlOn of actI:e, de�pIte t�e. fact that stand�rdization procedures seek to strictly limit
The analytic objew how what is being said relates expenences and the mterVIewer s mput and restnct the respondent's range of interpretive
talk but to sho ied in the circumstances at hand. toViethe inter�iew actions (see, for example, Houtkoop-Steenstra, 2000; Maynard et al., 2002;
live� being stud analysis must be every bit as rigorous asgthetheanalysIs of
win Schaeffer and Maynard, 2002).
as active means strued interview data. Analyzing "a�t�ve" interview data Second, by �pecifY�r:g . the vision of an active interview, we are not simply
conventionally cone, methodical procedures, and sensltlv . ity to both process offenng. an obh�ue cnticlsm of standardized interviewing methods. In many
require disciplin wa!s, sta.nda�dlzed survey practitioners are intimately aware of the details
and substance. findings from interview data is itself an analyt�cally active of mtervlew. mteraction (see Converse and Schuman, 1974, for example).
Writing up s .as a matter of lettm g the �ata They deal With .
them as mtegral parts of the craft of interviewing. But their
enterprise. Ratherselvthan viewing the procesmca . documents the meamng ­ vessel-of-answers model provides the basis for problems and methodological
"speak for them. Wites," the researcher emp refellyrenc e to records of talknmg , �he cha�lenges t�at are endemic to that model, not to interviewing in general. By
making process ve activ h ample illustrationwhiandch respo�den ts pr?d�ce mea callmg att.entlOn. to the acti:rity inherent in interviewing, we are simply offering
complex discursi goal ities through r lmkages and a competmg eplsten:ologlcal model of the interview in order to expand our
are described. Thevely conisstituto tedexplicate how rview envtheI
mea nmg s,
��nt. Reports ways of understandmg how we come to know about inner lives and social
horizons, are actisummarize and orga within the inte rview iron iCIpants have wor�ds. This model, of course, is replete with its own challenges.
do not so much -construct" participants nize what inte thepart F1�all�, by treating th� inter:riev: as active, we are not saying that "anything
said, as they "dets of narratives of lived exp ' talk to show reader both the goes wI:h r:spec: to mter.vlewmg technique and analysis. The concept
haws and the wha
erience. of the a:tIve mtervlew provides us with an alternate way of construing the
productlOn and collection of information that demands its own set of method-
010gical and analytic principles and guidelines related in the first instance to
the d�stinction between the whats and haws of the interview process (see
Gu�num .and. Hol�tein, 1997; Silverman, 2001). Indeed, acknowledging that
the mtervlew ls actlve may render interview research even more complicated
. approaches already hold it to be. The active view widens
than conventIOnal
1 57
1 56
 THE ACTIVE INTERVIE W
JAMES A. HOLSTEIN A N D JABER F. G UBRIUM ,ill INTERVIEWS

the anal tic purview of interview researchized to consider a. -:ider arra y of Atkinson, Robert (2002) "The Life Story Interview," pp. 121-40 in Handbook of Interview

uestio!s than are thenobail ralIs

iwick of standard cataorlognatuwha tIc app
s : , an.
Id
oa
111
ches Research, ed. Jaber F. Gubrium and James A. Holstein. Thousand Oaks, CA: Sage.

f. he rese archer can longer activ be content to t was � w�at Baker, Carolyn D. (2002) "Ethnomethodological Analyses of Interviews," pp. 777-95
.
m terVlew. The challenge of the
e inter view is to care fully cons Ider
. f-lal In' formatlOn
in Handbook of Interview Research, ed. Jaber F. Gubrium and James A. Holstein.

, where, when, and by who

is said in relation totohowwha � exp�nen Thousand Oaks, CA: Sage.

IS. conv. eyed, and much wide t end. Construing theegat intervIew
. as actIv. .e, then, Berger, Peter L. and Thomas Luckmann (1967) The Social Construction of Reality. New

rovides us with a view socier, ty,morwee richl y vari ed field f mqm .0 ' ry th�n York: Doubleday.
Blumer, Herbert (1969) Symbolic Interactionism. New York: Prentice Hall.
�ver before. In an inter must attend to both nen th � wh �ts and haws Briggs, Charles (1986) Learning How to Ask: A Sociolinguistic Appraisal of the Role of the
of interview practiceuced if we wish to documeny scent how expe tlal data are Interviewer in Social Science Research. Cambridge: Cambridge University Press,.

assembled and prod on the contemporar e (see Rose , 1998 ). Cicourel, Aaron V. (1964) Method and Measurement in Sociology. New York: Free
Press.
Cicourel, Aaron V. (1974) Theory and Method in a Study ofArgentine Fertility. New York:
Wiley.
Converse, Jean M. and Howard Schuman (1974) Conversations � t Random: Survey
Research as Interviewers See It. New York: Wiley.
DeVault, Marjorie (1990) "Talking and Listening from Women's Standpoint: Feminist
Strategies for Interviewing and Analysis," Social Problems, 37: 96-117.
Douglas, Jack D. (1985) Creati'ue Interviewing. Beverly Hills, CA: Sage.
Fontana, Andrea and James H. Frey (2000) "Interviewing: The Art of Science,"
pp. 361-76 in Handbook of Qualitative Research, ed. Norman K. Denzin and Yvonna
S. Lincoln. Newbury Park, CA: Sage.
Foucault, Michel (1979) Discipline and Punish. New York: Vintage.
Fowler, Floyd J. and Thomas W. Mangione (1990) Standardized Survey Interviewing.
Newbury Park, CA: Sage.
Garfinkel, Harold (1967) Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice
Hall.
Gorden, Raymond L. (1 987) Interviewing: Strategy, Techniques, and Tactics. Homewood,
IL: Dorsey.
Gubrium, Jaber F. (1988) Analyzing Field Reality. Beverly Hills, CA: Sage.
Gubrium, Jaber F. (1989) "Local Cultures and Service Policy," pp. 94--112 in The Politics
of Field Research, ed. Jaber F. Gubrium and David Silverman. London: Sage.
Gubrium, Jaber F. (1993) Speaking of Life: Horizons ofMeaning for Nursing Horne Residents.
Hawthorne, NY: Aldine de Gruyter.
Gubrium, Jaber F. and James A. Holstein (1994) "Analyzing Talk and Interaction,"
pp. 1 73-88 in Qualitative Methods in Aging Research, ed. J. Gubrium and A. Sankar.
Newbury Park, CA: Sage.
Gubrium, Jaber F. and James A. Holstein (1995) "Biographical Work and New
Ethnography," pp. 45-58 in The Narrative Study of Lives, Vol. 3, ed. Ruthellen
Josselson and Amia Lieblich. Newbury Park, CA: Sage.
Gubrium, Jaber F. and James A. Holstein (1997) The New Language of Qualitative Method.
New York: Oxford University Press.
Gubrium, Jaber F. and James A. Holstein (2000) "Analyzing Interpretive Practice,"
pp. 487-508 in Handbook of Qualitative Research, Second Edition, ed. Norman K.
Denzin and Yvonna S. Lincoln. Thousand Oaks, CA: Sage.
R E F E R EN CES Gubrium, Jaber F. and James A. Holstein (2002a) "From the Individual Interview
to the Interview Society," pp. 3-32 in Handbook of Interview Research, ed. Jaber F.
and Analysis: London: Sal?e.
Al U utari Pertti (1995 ) Qualitative Methods .
t�
Gubrium and James A. Holstein. Thousand Oaks, CA: Sage.
ide David L. and John M. John son (1994 ) "Critena for Assessl�g InterpretIVe
:
A Gubrium, Jaber F. and James A. Holstein (eds.) (2002b) Handbook of Interview Research.
\ lidity in Qualitative Research," pp. 485-99
in Handbook of QualitatIVe Research, Thousand Oaks, CA: Sage.
ln. Newbury Park, CA: Sage.
ed. Norman K. Denzin and Yvonna S. Linco
159
1 58
JAMES A. HOLSTEIN AND JABER F . G U BRIUM , INTERVIEWS
V
THE ACTIVE INTER IEW

Gubrium, Jabcr F. and James A Holstein (2003) Postmodern Interviewing. Thousand Rogers, Carl R (1945) "The Non-Directive Meth
Oaks, CA: Sage. od as a Technique for Social Research,"
Amencan Journal of Sociology, 50: 279-83.
Harding, Sandra (ed.) (1987) Feminism and Methodology. Bloomington: Indiana Rose, Nikolas (1998 ) Inventing Our Selves. Cam
University Press. bridge: Cambridge University Press.
Sacks, Harvey, E:na�uel Schcgloff, and Gail Jeffe
Heritage, John (1984) Garfinkel and Ethnomethodology. Cambridge: Polity. rson (1974) "A Simplest Systematics
for the OrgamzatIon of Turn-Taking in Conversat
Holstein, James A (1993) Court-Ordered Insanity: Interpretive Practice and Involuntary , ion," Language, 50: 696-735.
Schac ffer, Nora Cate and Douglas W. Maynard (2002) "Stan
Commitment. Hawthorne, NY: Aldine de Gruyter. dardization and Interaction
in the S�rvey Interview," pp. 577-601 in Hand
Holstein, James A and Jaber F. Gubrium (1994) "Phenomenology, Ethnomethodology, book of Interview Research, ed. Jaber
F. Gubnum and James A Holstein. Thousand Oaks
and Interpretive Practice," pp. 262-72 in Handbook of Qualitative Research, ed. Norman , CA: Sage.
Schutz, Alfred (1967) The Phenomenology of the Social
K. Denzin and Yvonna Lincoln. Newbury Park, CA: Sage. World. Evanston, IL: Northwestern
University Press.
Holstein, James A and Jaber F. Gubrium (1995) The Active Interview. Thousand Oaks, S�lverman, David (2001 ) Interpreting Qualitative Data,
CA: Sage. Second Edition. London: Sage.
. S1l1g1eton, Royce A and Bruce C. Straits (2002) "Surv
.
Holstein, James A and Jaber F. Gubrium (2000a) Constructmg the LIfe Course. Walnut ey Interviewing," pp. 59-82 in
Handbook of InterView Research, ed. Jaber
Creek, CA: AltaMira. F. Gubrium and James A Holstein.
Thousand Oaks, CA: Sage.
Holstein, James A and Jaber F. Gubrium (2000b) The Self We Live By: Narrative Identity Smith, Dorothy E. (1 �87) The Everyday World as Probl
ematic: A Fcr:linist Sociology. Boston:
in a Postmodern World. New York: Oxford University Press.
Holstein, James A and Jaber F. Gubrium (2004) "Context: Working It Up, Down, and
Northeastcrn UmverSlty Press.
Terkel, Studs (1972) Working. New York: Avon.
Across," in Qualitative Research Practice, ed. Clive Seale, Giampietro Gobo, Jaber F. Warren, Carol AB. (2002) "Qualitative Inter
Gubrium, and David Silverman. London: Sagc. Viewing," pp. 83 -102 in Handbook
of InterVIew Research ed. Jabcr F. Gubrium
Holstein, James A and William G. Staples (1992) "Producing Evaluative Knowledge: and James A Holstein. Thousand Oaks
CA: Sage. '
The Interactional Bases of Social Science Findings," Sociological Inquiry, 62: 11-35.
IIoutkoop-Steenstra, Hanneke (2000) Interaction and the Standardized Survey Interview.
Cambridge: Cambridge University Press.
Hyman, Herbert H, William J. Cobb, Jacob J. Feldman, Clyde W : Har:, and C� arles
H. Stember (1 975) Interviewing in Social Research. Chicago: Umversity of ChIcago
Press.
Johnson, John M. (2002) "In-Depth Interviewing," pp. 121-40 in Handbook of Interview
Research, ed. Jaber F. Gubrium and James A Holstein. Thousand Oaks, CA: Sage.
Kirk, Jerome and Marc L. Miller (1986) Reliability and Validity in Qualitative Research.
Beverly Hills, CA: Sage.
Kvale Steinar (1996) InterViews. London: Sage.
Macc�by, Eleanore E. and Nathan Maccoby (1954) "The Interview: A Tool of Social
. .
Science," pp. 449-87 in Handbook of Social Psychology, ed. Gardner Lmdzey. Readmg,
MA: Addison-Wesley.
Madge, John (1965) The Tools of Social Science. Garden City, NY: Anchor Books.
Manning, Peter L. (1967) "Problems in Interpreting Interview Data," Socwlogy and SOCIal .
Research, 51: 301-16.
Maynard, Douglas W., H. Houtkoop-Steenstra, J. van der Zouwen, and N.C. Schaeffer
(eds.) (2002) Standardization and Tacit Knowledge: Interactwn and Practlce m the Survey
Interview. New York: Wiley.
Mishler, Elliot G. (1986) Research Interviewing. Cambridge, MA: Harvard University
Press.
Mishler, Elliot G. (1991) "Representing Discourse: The Rhetoric of Transcription,"
Journal of Narrative and Life H istory, 1 : 255-80.
Pollner Melvin (1987) Mundane Reason. Cambridge: Cambridge Univcrsity Press.
Pool, I ;hiel de Sola (1957) "A Critique of the Twentieth Anniversary Issue," Public
Opinion Quarterly, 21: 190-8.
Reinharz, Shulamit (1992) Feminist Methods of Social Research. New York: Oxford
University Press.
Riessman, Catherine Kohler (1993) Narrative Analysis. Newbury Park, CA: Sage.

1 60
161
 MEMBERSHIP CATEGORIZATION AND INTERVIEW ACCOUNTS

rearranges them into a different formation. The words spoken by the

respondents and the ideas they are heard to represent are 'the data'. This is a
common-sense view of interviewing in which the interviewer, and later the
analyst, investigate 'interiors' (states of mind) or 'exteriors' (descriptions of
M em be rs hip social settings) through a representational view of language.
TALK AS SOCIAL ACTION
Carolyn Baker
Another way of approaching the analysis of interview data brackets this
common-sense perspective, and brings into play a different set of assump­
tions about language and social action. Drawing on earlier work in which I
analysed interview talk between an adult researcher and adolescent subjects
(see Baker, 1984; Silverman, 1993), I will explicate further how the use of
d methods of dataa greageneration in 'membership categorization devices' is a key to treating interview data differ­
IntervIews are amonge the mos. ' t widdelh use Wh ile t deal has ently. In this perspective, 0) interviewing is understood as an interactional
the soci�� :����:;e�ale�����7h:�roc�dI�t;:' �'for200ge2)�erating
.
�

�
.

such ata (hO event in which members draw on their cultural knowledge, including their
b een wn t . how to re1a te) rather less attention has been gIven un�1 knowledge about how members of categories routinely speak; (2) questions
ns,
to ask questIoently
relatively rec to the �nal. YS::� oaf:alysis of he inte data as
focus
s on analysis, a much
uch data� A focuractio are a central part of the data and cannot be viewed as neutral invitations
to speak - rather, they shape how and as a member of which categories the
. s expe�t1se in ese Slgr
on the researcher'tIOn . ifica ntly
nal
how interviewing may be respondents should speak; (3) interview responses are treated as accounts
as in the genera pursue of It, ch ang : . more than reports - that is, they are understood as the work of accounting
understood and d within the socIal SClences. by a member of a category for activities attached to that category. This account­
ing work is the core of the analysis of data. In this accounting work, we
look for the use of membership categorization devices by the interviewer and
I NV ES TIG AT I N G I NT ER
IOR S AN D respondent, and show how both are involved in the generation of versions
nce perspectives,ofthesucrele
vention.al social sciedata vant �esear�her of social reality built around categories and activities. Further, in the work
From many �onthe and criteria cess at mterv1ew­ done with categories and activities, we see the local production in each case
expertise is m h magetttertmgs as0f the here was good whethewhe ther the of versions of a moral order.
ing include suc ed a lot a�h:thh:� ;the 'rap por t',
Y talkred. Allabosuch criteria of suchow
ut, r and It is helpful to understand that from this perspective the process of inter­
respondents talkwhat the mte : rviewer was aftesting information of some cess
. viewing is better described not as data 'collection', but rather as data 'making'
they d1VU' I ged there is pre-exi sort or data 'generation'. The analysis of interview data from talk or transcripts
a hon that f e of the talk is organized not to locate interior beliefs or knowledges or to seek
�:�r :l���:, ::e t�I:des, l��owI;dt�e�;���;������r �i:�l��S ���:��:, :rl::� actual descriptions of social settings. Rather it is organized to identify the
�;���re:��� ;: ��:/th�fh: resp�.ndent sfoe:::si�feer��' �::��::���l7ta�: speakers' methods of using categories and activities in accounts. This is a
e ;:l!�;:: ��o��n of special connection between round-about way (but the only one possible) of identifying cultural knowledge
truthh:llY' to proW�� � and logic in use. Cultural knowledge is audible and visible in how people
f affa.lTs. From re1sp1�ndeni t an intei rview can be found to be good or not,
?mte account to one another, whatever might be inside their heads. This approach
rVIewer and ' draws also on Sacks's introduction of the possibilities of using conversational
suc sful �r r:0t.wh' ch follows from this conventional perspective typically
ces data to do sociology, and on Silverman's many analyses of interview data.
The ana ys�� ;e cor tent 0f what is said by the respondent. This perspec­ It sets up the interviewer and respondent as ordinary competent members
seeks 'themes mw dt �ta m1g: �t b c t db the hrase 'from thought through of the culture and the analyst as post hoc ethnomethodologist, looking for
tive on intervie mes That IS, tl�e �;n�:�ts J the ;espondents' thoughts (be:iefs the social-organizational work being done by interviewer and respondent.
language to the ed m: the m dium of Ian ua e (the interviewer's task IS to The speakers are viewed as competent observer-analysts of the interaction
etc.) are exp�essexpressIOn
. ) an� . then this co �te� is reth ema tize d by the analyst, they are involved in. It is their artful use of talk we look for in analysis. In this
enc our age thIS . them around and approach, it is not interviews that are good or bad, successful or unsuccessful,
who typically chunks the data, categonzes them, moves
1 63
CAROLYN BAKER INTERVIEWS
MEM BER SHI P CAT EGO
,{

RIZA TION AND INTE RVI

EW AGG O NTS J
in themselves. The criterion of success is the ability of the analyst to explicate I N TE RV IEW TA LK
the routine grounds of the work that interviewer and respondent do together AS AC CO U NT IN G

to assign sense and meaning to the interiors or exteriors they talk about. The first seg nt from interview talk to be studIe.
interview thame . e I. �1sht. utl?Onher1 epracomes from an
recorded for trestooeark chplapu
ce as � art �f :o��:�
. erv�Iew extctiracce but was
the opening turns of an aud. rpo es. h wm g mt
d. e� parent-teacher interview (Batkershoand ws
Keogh, 995) . Rig")ht from tiIo1e-rebcor
M E MBERS' A NALYTIC R ESOURCES
]
Members have analytic resources that they put to work as they engage in work with their Own and the 0thers eglnnlng' participants Im ' med'lately go to
any kind of talk, including interview talk. One of those resources is the use encounter (see the Appendix to thOIS 'vomeIum mbership categ. or.les '
e for transcnptIOn rekeyI eva nt to this
of membership categorization devices as introduced by Sacks (1992). Talk, ).
further, is not simply expressive of interior states or contents. Talk is social Tea cher (T) : Elle n
action: people achieve identities, realities, social order and social relation­ Stu den t: Don na
ships through talk. How people describe things and how they reason about Parent(s): Mo the r (M) and Fat her (F)
them are pragmatic selections from a range of possibilities. Even 'simple'
describing is always a social and moral activity (Jayyusi, 1984, 1991; Scheglo£( T: Ok all righ t we 'll just forg
I hate tape rec ord ers ! (hh) �
et it I sho uld cov er it � or
'
som .
eth ing
1988) turning on category identifications. Imagine the differences in being M ister J ay ' s clas s um fou
Rig ht um D onn a um I Jus
t too k ove r
approached to speak as 'a mother of three' and being approached to speak r weeks ago so I don 't
� ��� �
rea ll k
��� :������:� � �:;� �:� �
'
as 'a professor'. Both may be correct identifiers, but the selection made calls d
a ' had a quick I O k at her w
rk e
on very different domains of knowledge and reason. To account for one­
e r
the first part of the year
h
� �� ��� ��� ���� ��� �
'
S t e
' e in
self 'as a mother' calls into play other related categories (such as 'children') um , pieces of work wh ich
and h r y
was a poetry ora l?. and
e i wo
and activities or properties associated with those categories (e.g. 'needs' of (2.0 ) a nov el in ano the r for
�.
m that was put lng par t
a um a nov el

children, caringness, guilt). Speaking 'as a professor' invokes a different set into ano the r sty le of wri tin
Now
of the nov el
(1 . ?)
�? �� � ���� ��
2.
of category relevances and activities (e.g. 'students', teaching, theorizing or ass Don na's

���:� :
a littl e bit distra cte d? ofte
. dow h
researching). When we are asked to speak in some situation, as in an interview h t she s its wit h, tho ugh
0, s h e d oes do all her
she does giv e in c s ��; :� �� :�:
h

for example, we mobilize the resources of available membership categoriza­ work ' um I ' m (1 . 0) wou Id
e h
.
you II ke
tion devices. These devices are collections of categories and associated to - do yo
� ;�� :�
k ' h Donna at hom e wit
h her sch ool wo rk at all?
+ +
do yOU se
activities (such as mother father children nuclear family). 2 F: No t rea lly no=
a r?

An anecdote will make more striking the point about members' uses of 3 M: =(W e very rarely) see
membership categorizations to assign social identities and achieve social 4 F : the y gen era lly d isap
her sch ool wo rk

order. Recently I observed the following scene. A woman, a young male child h om ew ork and um =
pea r off to the ir bed roo
ms wit h the I'r
holding a screwdriver, and a man entered a bakery. The woman behind the 5 T: =Ye:es (2. 0) We l l um
counter greeted them and then addressed the child as follows: 'Have you 6
;�.
F: We don 't see mu ch of
�
(it)
been helping Daddy?' There was a pause that I distinctly remember because,
7 et me see yes I d id n 't mar
k this this was all Mr Jay
as an observer-analyst of the scene, I had time to consider the gender assump­ IS IS a sum mary, they had
to sum ma rize um this (1
's (1 . 0)
. 0) um Iet , s
tion being made: that Daddy was being helped rather than the mother. The see whe re her , m istakes
.
see m to lie (3 0) Oh I't see
woman customer then spoke: 'This is not Daddy.' What resources had the (3 . 0) Why d'd I S he only get fou r and
'
ms a l ngh t .
a hal f for tha t. Hm mm .
woman behind the counter used to generate her 'mistake'? Should we take In the teache g first turn, she de 'b;� er
it that the woman customer is the child's mother? Is this Mommy? How do we and accountsr'sforlon pos sib le sho rtc om in; � w rk �' n consId. e:able detail
:� .a wor� pnor to turnmg the talk
hear her that way? over to the parents or Donna . The turn �com bmes information seg ments (how
Category incumbencies are 'made to happen', are produced (and sometimes Donna is doing, how she is in cla
ss) w'1 th a
corrected) in talk. We routinely and pre-reflectively use membership cate­ is herself doing in the interview) The �����llogatl� segments (how the teacher
gorization devices to organize our characterizations of what we see or hear. the category of 'teacher' as one who ' h st-�andOn can be seen as depicting
Membership categorization is a pervasive resource for sense-making through of the student; can read other teache::' �0tatIO. and long-term kn owledge
utterances. Tracing members' use of these categories and devices in any observe at is going on where l' the c1assroons and make sense of them;
settings, including interview settings, is a means of showing how identities, friends ins wh m; s who the student's
;ows wl� �theta�ents wantkntoowkno
social relationships and even institutions are produced. of the turnclawissthareth' eaninfd korm ation :rlat s le IS not Donna's lonw.g-tThime pre facing
e teacher
1 64
1 65
CAROLYN BAKER INTERVIEWS
MEM BER SHI P CATEGO
RIZATI ON AND INT ERV
IEW ACC OUN TS

and then turnmg . dlr. �ctly to onn , marks is also a characterization in her who 'disappear off . . . with their homework'. It
talk of the category pa�en � ��se who want to know about Donna's wait to get at homework, she disappears offcould be as if Donna cannot
academic achievement m t�� �� s\ instance and want to know about her parents do nother with her. The teachersois fas t, and that is why the
behaviour in class. The teacI er s tl t t rn can be heard as an elaborate version this scenic descriseeptiitonor ofwohorkrne cal led upon to recognize
of 'who I am/what I know /w1�at :y �elevancies are, and who you are/what We now have at least two Dolife .
you know /. what your relevancles. ' are, and what we want of each other'; a way Donna who sits with the girls at nn the
as produced in this talk: distracted school
0
of connectmg th. e tw . inshtutlOna 1 categorl'es, teacher and parents. D onna, in class), and keen home Donna whbac o
k (but nevertheless does do her work
dis appears off to do homework Both
the subject of the talk, .IS p�esent but Sl'Ien.t· The teacher's work here can be seen are moral constructions of Donna at the sam
' colounng
as idenh'fymg, m
. , and. connec tmg the two institutional categories of structions of the speakers themselves as observe tim e that they are moral con­
arent and
PconnectlOns teacher. In Sllverma
. between cultural partlculars. � , s (199::)
' 114) terms , she is assembl'mg This analysis suggests a way of beginning aantmeparents and teacher.
device analysis interview talk. We have in this mbership categorization
� .

The teacher also calls . on unnamed but alluded-to student categones . and her schoolofach ' , This topic is inttalrod k a topic which is 'Donna
in scemc' descnp' flOns such" as 'down the back there, with urn the girls that elaborated through varieviouems ent posed descriptors of Douce d explicitly, then
she sits with'. These descnbmgs are forms of social action . done through talk descriptors can be seen as 'acpro nn a's
ms of the notion of 'caacttegiviory
ties. These
(d. ScheglofC 1988). activities' which imply membtiversitiehips' inin ter ries. In this case, the descri-boptounrsd
At the end of the fl..rst turn th t cher changes the topic to what the parents do not all get 'bound' to the category Docatnnego
do at horne concermng school:'��k. The teacher's question to the parents, since the speakers go on to delve furthea,r no int
r even 'attached' very firmly,
'would you like to do you k 'th Donna at horne with her schoolwork descriptions of Donna, then, are candidate descrio Donna's activities. These
at all? do you see It: at all orV;;· ?�eg��s like a perspective display' series (d. not turn out to be 'attached' to Donna in the end ptions, which may or may
Maynard, 1991) asking ,
for the cl'lents' view of matters but then mtroduces
.
state or imply Donna's membership in possible cat. The descriptions variously
what looks like an mformatlo � que. �flO�It" This is information seek'mg, but student, keen dent, and so on. The talk is a dis egories such as distracted
0
in terms, f In' teraction in the. mterview IS more than that. It is a further of how such stastutem ent s and play of speakers' knowledge
elaboratIon fthe category 'parents, done by adding to 'parenting', ,as a1ready
0
im pli cat ion s mi gh t
use what might be termed SOcial-structural andbe heard by the other: they
established, the category-related ttr 'b te of possibly working WIth Donna certainly at least inter-institutional knowledge tha cultural knowledge. It is
or seeing what she �oes. at h e ; S�:ething else. The question gives th�ee descriptions of Donna. It is in this sense that thet they draw on to do these
options of descendmg mvo�:nent for what parents might do: work wIth insight into what they know. What they know is talk provid es us with an
Donna 's schoolwork (at all ever.? somefImes.?), see Donna's schoolwork (at how to do accounts in precisely this inter-instituhow to do deSCriptions and
all ever? somet'lmes.?), or?. (hearably, something even less.?). that their descriptions of Donna are reflexively tional setting. Remember
:=:

'they disappear off to their bedrooms', placed jusdescriptions of themselves:

:=:

is effectively a scenic description of a parent wh t where it was in the talk,

i NTERVIEW TALK AS CATEGORY ELABORATIO N with homework, not by his own choice. o could not po ssibly help
In this interview, then, there is much more than
Before the parents have had SI. gIe turn at the formal business of the of information or pe rsp ect ive s abo ut Do nn a
the asking for or getting
interview, the teacher has, using\er�( OWled?e of category memberships and itself is a site for displaying the cultural knowled an d he r work. The interview
related activities, presente� � ,comP�ex sOcIa1 landscape that connects the for oneself as petent parent or teacher. Th ges tha t can be used to account
" 0
interests, terntones and actIvIties f parents and teachers, The landscape is on the namingaofcom ese
sometimes merely alluding to category,alcat cultur knowledges turn
also a moraI one now m' several ' 1udl'ng the descending order f
respec ts, mc 0 or category-bounord act egory relations
parental help to Donna la, d ut as the teacher asks the parents to speak. What this recommendivistiefors. the analysis ofinterview
As member-analysts � t�IS scene, the. parents can take it that the teacher
0 of the category knowledges at work in the tal data is the identification
wants to know what happens at horne an d where it fits in her descending need not be contained in elaborated turns, bu k of the participants. These
order. ,Not rea11y no' and. 'we very
. rare ly see her schoolwork' are heavily shorthand - with the same effects. As seen t can be sketched spoken in
to t
mitigated ans,:"er� �e mfor uestion The sense these turns produce call on and convey a great deal of cultural knowabove, a single utterance can
_

is that somethmg,mter eres WI�:��: �gitimat� parental task, and that is th�t placement in talk ledge through its design and
'they generally dIsappear off to their bedrooms with their homework'. TIm The pa
is an artfully constructed account. Who 'they' is is not clear, but it is more than with accourenntit-tngeaconherbotinth sid
erviews from which this extract Was drawn are thi
They are 'inter-views' in the original sense ck
just Donna. Donna has now been pI aced in an unnamed category of people 'entrevoir' - 'see each other' -pes.erh aps more than are most research interviewofs.
1 66
167
CAROLYN BAKER '" INTERVIEWS
MEM BER SHI P CAT EGO
RIZ ATI ON AND INT ERV

The talk is not symmetncal . ut questionin is done by both parties, and there the people po ed as 'adolescents' in the int
IEW ACC OUN TS

is often a pract'1caI outcome f r;"olutions

0
. to. mta? 11theThform1. nterview
, or advice towards which categorizationsitanion erview d membership
estions. Especiuse
the participants contmue is often conducted or orgamze
by at least two part�es, an� th:ee : e the student is present and both student
. d category-bound oralycatsisegoinry-animswplieriedngacttheiviqu
ties attached to eacallh yofinthethelifeir sta
use of
and parent(s) are mvolved m ra�si�g toP ics or responding to them. In one that the interviewer had presented as giv ges
accounts which displayed their cultural knowen,ledge
<..

. WIt. h the t-acher' father and stude11t present, the interview became the respondents produced
intervIew of adolescence. e this one step further, they disabplaouyed t adult constructions
an extended speech by t�le father with the teacher positioned as aud'le�ce of culture as fraIfmewed tak through the adult lens of life-stage folthekirpsy knowledge
and the student mostly sIle�t. Th�se arent-teacher interviews have famIly This is the particular up
e

cul chology.
resemblances
. towhmany
consultatIOns
healt , clinical �nd
ere problems concernmg . other
0
institutional encounters and
ne of the parties to the. talk (the in the interview, on the intturerv al knowledge they activated in order to tak
iewer's terms. It is cultural knowledge inethrpaeert
0
client) are discussed. The �pproac.h t n lysis represented here is mterested s enses: first, it is knowledge abo ut the culture; second, it is not specific to
m. the pragmatics of the mteractIOna ts:tti g . How .people do things with
individual respondents but reappear
words, and what they d WI'th them 0 . , are the�focus of mterest it demonstrated a 'Successful' readins ing ofdifthe ferent people's accounts; and third,
particular interview situation:
. how does this son want me to speak? Which
of characterizinper membership is in play here?of the many possible ways
Letting go theg my
of privileged insighpret intsum
o
ption that (good ) interviews give us some kin
wh ple really think, believe or do is the firsdt
step to seeing interview data atas peo
M E MBERSHI P CATEGORIZATION WORK I N RESEARC H
I NTERVI EWS

. . are those which are conducted for research whatever is the topic the interview the production of situated 'accountings-for'
Another category of mterviews
purposes on1y, Wh I'ch. would. not have taken place had the research analysing how people talk to one anerothpreer,senonts.e In Silverman's terms, '[bJy
been undertaken. Unhke t�e .mterview described above, research mterviews . . not a cultural universe and its content of moral assumis directly gaining access to
ptions' (1993: 1 08) .
are designed
. andt tconducte . m order that respondents will speak about some
topic of mteres the. researcher. rn such m' terviews a more asymmetrical
0
organization of talk IS usually see�, WI'th the interviewer asking questIons . ME MB ER SH IP CA
lO GIC S
TE G O RIZ AT ION AN
D CU LT U RA L
or making probes, while the mterviewee talks at more length to supply the
information s�ught A ve ry lar ge
Such intervIews are typica . Ily conducted under the traditional social science
stand wh�. ch as'ks res pondents the purpose ofpro
fin
portion of research interviews would be con
din cific information, perspectivesduorctebeldieffors.
or on their external
.
world as
to reveal, desc r'be
the y know it.
1 , report on their interiors
The respondent is positIone
. . d
Such interviews are typg ouicatllysomchae spe
racterized by a very asymmetrical organiza­
.
essentially as a. WItness f h1' s or her own. m. ten' ors and exterior circumstances
0
tion of talk, in which the interview er asks the questions but talks much less
who gives testImony t h1' s or her expenence of events. In this mode of dOl1. 1g than the respondent. In such circumsta
0
social science, the stlcky problems fbIa' �(On the interviewer's part) and truth-
0
part the talk done by the respondentnce to
s, the material for analysis is in large
make available to the interviewer
telling (on the witness's part) come to t � fore' whatever information is asked for.
In a study
These problems Were encountered In an alyses of research interviews
conducted WIt. h 'young adoIescen ts' (Baker, 1 984) . These interviews capture the senunsederantak
d
en of a school's 'welfare system', I Was intereste
log ic behind the system. The system in question d to
saturated with category talk m. the o of owerful assumptIons . awere bout an elaborate system of teache
behaviour. Tickets were physir-acalssilygnissedue'tic kets' to students for good or powaors
activities attaching. to th� cate��ry ,a� �:cent (see Silverman, 1993: 90-114 good achievement or behaviour, or whited for to students, either yellow ones for
for extracts and discus�lOn). 1ese as�umptions were lodged right inside My knowledge of the system came from mi sbehaviour of some kin
the interviewer's questIOns, for �xamr:I ' hen do you think you'll be an
adult?' The resulting talk by �� � ter��:�:s to uestions such as this was occasions and attending a staff meeting in whvisiting the school on sevd.eral
rife with. talk about resp�nsI I. I.;�� �� end of �arefree living, and other being discussed at considerable length (see ich changes to the rules were
the staff meeting talk). Baker (1997) for an analysis of
cultural Icons of conve�tlOnaI a hood. The way out of this incestuous A talk s later arranged with the Chair ofthe
. b�twee� 1' nterviewer cat egon'es and interviewee categories was tok
relatIOn the Chairwa to give his view of the system, a 'perspWe lfare Comm
treat the mterview data a.s d1S' t:lays of membership categonzatlOn . war ective display'.ittee. I invited
,

by interviewees as well as mtervIew r. What this resulted in was in the

� I ar m. dividuals at all, but about hoW end
not about 'adolescents' or these partlcu I: Um . Wh at's wh at's you
2 C : [It's 9@ill. It's um .
r � ( ) of of the sys tem
(1 .0) the good kids it it kin
[is it
d of recogn ize s what
1 68

1 69
 MEMBERSHIP CATEGORIZATION AND INTERVIEW ACCD iJNTS
CAR OLY N BAK ER
INTERVIEWS
As the account proceeds there are som . . made to the problem
e re�lslOns
the ir the ir efforts ( ) the
y get. The kid s wh o it is
hard for (1 .0) wh o .
and a form of response to the probl em IS descnbed. The Ch . . alr. remarks that
don 't get the yel low (.)
tickets um ( ) kid s who jus
t plo d alo ng and
8 taff are encouraged to 'look for the lodd�r�', although It not clear what
18
the y usu ally don 't, mig
htn 't get a yel low ticket
for bec aus e the y
go up the system is supposed to be done differently lo 1° ers except to recognize them.
haven 't don e anythin g
extra great ( .) but they stil l
Here 'plodders' are produced as t�Fse:' �� s��dents any teacher co�ld recognize.
The category device now con�e t ff a d ��dents [good kIds, plodders,
't get a wh ite ticket for ten
get a tick et, don
bec aus e, if they don 't s tha t, you know,
y. Um (2.0) they're the one
weeks the y go up anywa
ff), if you look for the plo
dders who are teachers] the cast of characters so far a we can ��te that they are
connected through attributions of em athy and care. Desc�lbl �g plodders now
_

you try to enc our age (sta

__

ting as much effort

as 'probably putting in as much eff�;' �ffec;s � reorgamzatIon
wh o are probably put
jus t plo ddi ng along but of 'plodder'
_

you kno w, the best marks i n

are getting ,
in as the (.) peo ple who
the cla ss, w e try and enc
ourage tha t. character (earlier, twice they 'just pI° a °ng ) . alk about the. distribution of
0
tickets (giving or not giVing) seems t re. fer to an underlymg orgamzation
up to plus
I
.
plo dde rs as you say to get
I: t is now pos sib le for the
of sympathy. Whether or not staff are f-act encouraged to 'look out for
.

3
fou r?
c: No . They can sti l
l onl y, stHI onl y get to (.) to
neg ativ e er pos itive
Tha t's bas ica lly
plodders', in this interview atleastth·l�res�onse to the proplem brings forward
111

4
two. ((So und of paper rus
tlin g. 4.0)) Let 's see. (1 0 .0) a version of a duty of care that teac ers lave towards plodders. This duty
the state of the nat ion
, that' s, the issu es. (1 .5) of care was not attached to these stuo dents when they were (ear1"ler m the
mi nim ally in his firs t answer: 'It's answer) 'just plodding along'. At th t earller ' pomt
a . were described as
' they
w is rep ort ed dir ect ly and
The Chair's viefollows is an account of how the system is working well, andg not. getting tickets and this was represented as a natura1 consequence of
great' . Whatt. This account turns on the early production of two contrastin thelr (plodding) behaviour' although ' t was . acknowledged that the system is
1

how it is no students: the 'good kids', whom the system rewards, and the 'hard' for them. . of th� ticketing system,
categories of us a membership categorization device of the form [goed. od Far from being merely a 'report' about the op�r.ahon
'plodders' . Th] is produced, to which further categories could later be add ] this account displays a version of the local practIcal reasonmg that members
kids, ploddersivities attached to the main contrast pair [good kids, plodderso could use to describe the syst em. The account is. des' Igned to convey the
There are act ries are produced. The system works well for good kids whers speaker's recognition of the categorizations, .�ohvatI. �ns and morality that
as these categoorts recognized, but it does not work so well for plodd do attach to teachers' work with the sYstem. ' That IS, the SOCIal world of the school .
(the 'exterior' 0
being talked about here) IS assembled as a comp1ex f categones
have their effd along' and never get a yellow ticket because they do not ers . .
and motIvatIOns which produce a moraI order as well as a setting of practical
who 'just plotra great' . On the other hand, the Chair remarks, the plodd as reasoning and action.
anything 'ex ite tickets either and are 'probably putting as much effort in' y Each move in the interview serves to add. to. and e1aborate �n the categories
do not get wh. . getting . . . the best marks in the class' - a category who ma and activities proposed in the initi�1 desc�lptI on. S,:ch cat.egones and activities
the 'people . the same as the 'good kids' he began with. are woven into a set of relationshl �alu��, mcludmg justific�tions and
or may not beiewer has heard a problem being stated, and asks whether the­ evaluations, in the course of the f:l��. g . n e ect . a moral ordenng of this
The interv an acknow
le. This questioniceis [go
up to the top ofthethecatscaego aspect of the work of the school IS' bemg done whIle the Chair provides his
plodders can nothew get blem and of air's conriz ation dev higod kids, views.
ledgement ofperfecpro matches the Chws the intcerervn iew about how h up the
The interview continues:
plodders]. It dderstlycan get, and sho erviewee characerterize the ploddthe
ent eri ng int o
system the ploBoth interview ers I: I
problematic. in this respect, erandandtheintsystem as autonomously constrainingir 5 Could get a copy of that?

as ambitiousThis organization of empathy with the plodders confirms the 6 C: Yeah you can have that copy.

their ascent. social world of the school. 7 I: Thank you.

reality in the of the Chair's answer is as important as the contents of it. Therebe (1 .0)
I ��I-
The design
C' U m we' welfare policy that [principal] a d raft

good kids and plo dders; these may. Thore ma y not 8

� ; �:� ��;��� ,
may or may notptorearsllyusebed els
o e th t we t t . give you a copy of that too.

here in discussinichg the system pro duction I:

the same descatcriegories is part ew
9 Thanks.

of the two rality. Lookingofatantheaccbeg ount wh elaborChateairs 's responsecou, weld

wh at we (1 . 0)
��� � : �� �
call a local mosystem is at once 'great', it rew inning of the rts of 'good kids', but 10 r
�� � t �: �:��;
s differen to last year's (.) i n that after three
,

hear that the the other kids that just plod alo ards the effo s it would appear it i e a � ng etter (.) home, and then five white

'. From thi

is hard on 'all of students is benefiting and angmajority is not.
tickets they go on to n g Ive one and then another th re� (1 . 0)
th ey get to er another warning l etter , and from here to th ere IS only
_...

that a minority 1 71

1 70
&

MEMBERSHIP CATEGORIZATION AND INTERVIEW ACCOU�TS

CAR OLY N BAK ER 9 INTERVIEWS

fou r ) is goin g ( two in a matter of a couple of weeks if they're having a run-in with
. e on (. ) a n like this .
e level two to negative level
.

thre e so from (. ) negativ

to be, ver y quic k ( ) pro gre.ssio n if the y con tinu
them . So now we're trying to, looking at, if one teacher's givin g out
three or five white tickets we try to kind of (1 . 0) find out what the
I: problem § with the teacher as well, more so than (1 .0) just the student.
I:
11 Mm m

I
(3.0 ) 16 Do you cancel the ticket?
d from last 1 7 C: Well it can be yeah (mean) we can't cancel them but we Q£!n
In turn 10 thetur ChaIr. off. ers the . tOpl. C of how thet issysimp
tem was changedes
lied thr oug h a cription
ther ��t;f�:Y 0fs�et�den
cancel them we ask the teacher
year's. In this lown ano This add itio nal cate gor y imp lied
of descent to utetick � levetselsand warnsYn lett. ers sen t hom e is the poo rly beh ave d In turn 15, the Chair embellishes the point he made earlier about quick
throug h talk abo . the m1. ?t'Ia1 descrl'ption. Such students are ones descent, that it is 'only' another three tickets that drives the student down the
studen t w h 0 d'd 1 n ot app ear m
. lly, ' cont'mue on like this' (getting mo re and motomre levels. This was the change from last year's policy introduced in turn 10,
CO Id , hY pot het lca
wl10 U and who would, m' fact, have a quick progression to the bot and it is a change that he, and not the interviewer, has made accountable. In
white tickets) tem. A downward descent IS' the only. direction implied in the making it accountable in this way, he has underscored its status as a prob­
of the .leve. Is sysf the Chan. 's d escnp. t'lOn, although presumably not everyone lem. Retrospectively, it seems the interviewer's earlier appreciation of what
orgamzatlOn 0 atlVe doe s in fact trav el dow n to negative level four,h it means was not stated strongly enough; she was not grasping the moral
who get s to neg . . I eve I two des ign ed to catch and stop suc point.
s y st� her e said to be
and , �;:��:;d descent is accounted for by the p.osSibl.e ,activ.ity
the The Chair's turn in 15 can be seen as having three parts, beginning with
descents. Thestuden s m. ' t O " f they continue on l1ke thIS , WhiCh the elaboration of the speed of descent and its consequentiality: that is, serious
att�ched to � �:� �� t� ;K;d�nts persisting in their inapp�opri.ate trouble for those students. At the centre of the turn is a central fact that holds
asslgn� the ��en�halI�;socl oncludin g formula tion ' so from two to four IS gomgas the whole turn, and the whole logic, together: 'Because a kid can get to this
beh avI Our . e . . n 'f they continue on like this' is first treated
1
stage in the matter of a wee! you know, well kids hav..f.', of which more below.
be
to , . a ver y q U lck pro gre . SSIO In the latter part of the turn, a new category of problem actor in the system
a descr�ptl' On. anwed r,app ear.s soon after Cha to be heard as ation problem.
, firs t pro vId es a wea� is described: teachers who give out too many white tickets to one kid. This
reC lpient of the ir's des crip
The 111tervle. theasassessment made by the Chair . After a pause, the Chan
.
category is generated as another account of how a kid could get to this stage
agreement WIthk agam. t which point the interviewer provides a stronger so fast. In contrast to the kid 'continuing on like this' in turn 10, now we
starts to spea rm. ng 'tl�e upshot that progression downward will be quick: have teachers having run-ins with kids and needing counselling themselves.
agreement, retu The description of these teachers appears to be something of a trouble source
I:
for the Chair, given the pauses surrounding the delivery.
11 Mm m
Within the third part of this turn, the Chair begins by describing the
12 (3. 0)
c: ( [ )
Committee's solution ('and (.) what we also look at is now is') before naming
13
I: ( [Qu ick, yeah .
the problem ('three white tickets if it's within a (.) um (.) from one teacher').
14
att . er's appreciation
. terVlew
11e 111
What this does is to assign alertness to the Committee, another category of
. n 0 £ the Ch air' s nex ttu rn sug ges ts th
The d eSlg has not been strong enough, that she is still heanng . It' as a actors within the school, on whose behalf the Chair is accountably speak­
sho t
of the up her than as a problem. ing. In the course of this elaboration, a different version of the kid on the decline
description rat has been produced, one who is at least in part possibly a victim of run-ins with
teachers.
14 I' ( [Qu ick, yea h. . There are different possible upshots that could have been produced in
. �
w h·te
I
15 C' SO it's only an extra from neg
ativ e level two whi. ch IS t en
( ) d relation to this turn. For example, one could be that run-ins with kids should
�
a
a
l thr
not happen and that this problem in teacher-student relations is important
e and the n negativ e leve �;
ti kets , it's only another thre
:�:� can be (.) It s �p
to discuss further. Another upshot could be that the run-ins are taken for
they
(.) an r ano ther thre e afte r that and the kind of
(2.0 ) that 's whe re all

:���
n sort of stage so it's um
er cou nse llor and er (2.0 ) all
the work with the parents etce
tera granted but they should not skew the ticketing system. The second hearing
� .
ha to be don e ( ) fairly quic
kly. Bec aus e a kid can get
1.-
to thiS
(.!
stag e
is produced by the interviewer in her question 'do you cancel the ticket?' This
in the matter of a wee k, you
know, wel l kid s hav e.
,
and wha t we
. formulation by the interviewer seems to ignore the Chair's description of the
s a look at is now is ( ) .
that ( . .
) thre e whi te tickets If It s with
in a ( ) u m
Committee's counselling work and attends to the Committee's moral action
f� in relation to the practical outcome for the kid understood now as a member
che r mor e than
to kind of cou nse l the tea
. 0) from one teacher we try
' so one teacher has n't
got that effe ct on one kid , so one
. of the category 'victim of teacher run-in'.
��� ���; cillY som eon e on neg ativ e leve l ()
t etic o ne tea che r can put
1 73
1 72
CAROLYN BAKER INTERVIEWS
MEM BER SHIP CATEGO RIZA
TION ANO I NTE RVIE W AC
OUN TS t
It appears that the interviewer has by now heard the central problem as These categories are in a sense the sp eakers" uppets,, .
up .in. dif ferent ways and make beIlave m. vanpous wayswh Ich they can dre ss
actIvItIes ). These are powerful sta te:��ts about hat egory-associ. ated
.
'Because a kid can get to this stage in the matter of a week, you know, well kids have',
and not as the problem of teachers and run-ins. The Chair's turn is designed (cat
around this central point, first leading up to it, and then moving sideways, the social order might be arranged, w e e or not �It reacou ld be the case, how
producing an account for it. In producing the account for it, what is under­ duction
. lly
of plausible ver'sions, usm. g recogm�zable membersh'is.IpThe artful. pro-
scored is the moral issue of 'How can this happen in the matter of a week?', deVIces is a profou important form of cultural compet categonzation
backed up by the adjacent point that it has actually happened. What the Chair and attend to in thendlseyinte rvi ew acc oun ts are me mb ers
ence, What we hear
is presenting here is an Extreme Case Formulation (Pomerantz, 1986) whose together a world that is recogniza" bly famI'113r, ' me tl d for putting 10 S
' orderly and moral .
design and placement may achieve up to three things at this point in the
interview. First, it may be produced as a correction to the interviewer's fail­
ure to appreciate the matter of speeds of descent sufficiently. Second, it
underscores the alertness of the Committee to the complexities of the system,
the Committee's competence. Finally, it produces the kid/kids in question
as the ultimate subjects of the Committee's duty of care.
What we see here is the conversational product of the identity 'Chair
of the Welfare Committee', which is the speaking identity he was asked to
assume. The interviewer's uptakes and their absences appear to have been
oriented to by the Chair in his work of moral description and accounting.
By examining the membership categorization and other resources the Chair
draws on in his accounts, the cultural particulars he produces for this listener, REFER EN C ES
and by noticing how the interviewer's hearing of the talk itself evidences
membership categorization work, we are able to see how deeply interactional Baker, CD. (1 984) 'The search for adultness. Mem
this interview is, despite apparently minimal input from the interviewer. talk', Human Studies, 7: 301-23. ' bership' work 111. adolescent-adult
Baker, CD. (1 997) 'Ticketing rules: Cate oriz
school staff meeting', in S. Hester and P �glinat'IOn and moral ordenn. g in a
(e�s), �ultU1 e In ActIOn: Studies in
lysis. L anh am, MD . Ul11Ver �
Membership Categorization Ana
C O N C LUSIO N : ASSEM B LI N G POSSIBLE WORLDS
pp. 79-102. slty Press of America.
Baker, CD. (2002) 'Ethnomethodological anal�
In the preceding sections of this chapter I have demonstrated some procedures and J. Holstein, J. (eds), Handbook of InterVlew ses of 111' tervlew
.
s,, . J. Gubrium
for beginning a membership categorization device analysis of interview talk. esearch. Thousand Oaks, CA
'

pp. 777-96.
111

Essentially the search is for how participants in the interview make use of the : Sage.
Baker, C : D. and Keogh, J. (1 995) 'Accounting .
resources of membership categorization. The first step is to locate the central 111�ervIews', Human Studies, 18 (2): 263-300. for achievement 111 parent-teacher .

categories (of people, or places, or things) that underpin the talk, including Gubnum, J. and Holstein, J. (2002) 'From theIiind"
any standard relational pairs such as parents-teacher or contrast pairs such SOciety', in J. Gubrium and J. Holstein (ed"s) an l;l�ual 111. ter,vIe. w to the interview
as plodders and good kids. These categories are sometimes named and some­ Oaks, CA: Sage. pp. 2. d ofTnterVlew Research. Thousand
00

times implied through the 'activities' that are attached to them. A second Jayyusi, L. (1984) Cate,{or3-3 ion and the Moral 0 der. Lon
step is to work through the activities associated with each of the categories in Jayyusi, L. (199 1) 'Val�es izat
and mo . d?n: .Routledge & Kegan Paul.
order to fill out the attributions that are made to each of the categories. The ral ' u d eme n�-{ rCom ul1
in G. Button (ed.), Ethnomethodol�gy ;nd the uma � 1catlve praxi� as moral order',
University Press. 227-51 . n Clences. Cambndge: Cam
bridge
attributions that are hinted at are as important as any stated in so many words:
hinted-at categories or activities or connections between them indicate the Maynard, D. (199 1) pp. 'The perspective-disP laY sen' es and the deh. ver
subtlety and delicacy of much implicit membership categorization work. A news', in D. Boden and D H Z' a y of diagnostic
Ethnomethodology and Con� ; :;;:���� � (e�s), �alk and So�ial Structure: Studies in
+
third step is to look at the categories attributions connections that members s am ndge: Polity. pp
Pomerantz, A. (1 986) 'Extremeecase
ns.' A way oflegi' t'ImI'z111. g. clai
sa
produce (connections between 'cultural particulars'), to find the courses Studies, 9 (2/ 3): 21 9-99.
form u l atlo
Y" 1 64-92.
ms', Human
of social action that are implied: descriptions of how categories of actors do, Rapley: T. (2004) 'The open-ended interview',
could or should behave. D. SIlverman (ed in C' Seale, G. Gobo, J . Gubnu' m and
As in the examples I have presented here, when speakers 'do describing', Ryen, A . (2 002) 'Cros),ss-cQua
ultural interviewin ' .' n JLond
litative Research PractIce.
on:. Sage.
they assemble a social world in which their categories have a central place. Handbook of Interview Researc
h. Thousand �'a�s, C' A . sage
� u bnu m and J. Holstein (eds),
. pp. 335-54.
1 74
1 75
CAROLYN BAKE R !! INTER VIEWS

.
satIOn, VO1s I and II ed G. Jefferson. Oxford:
Sack (1992) Lec tures on Conver
< "

BlackwelL . 111. the social sciences I: Talk-in-interaction', IpRA

LX
5, 1 1-

'Des . tlOl1
cnp
Sc1leg 1 ff . (198 8) <

E A. . (1 / 2) .. 1-24'
0 ,

Papers in Pragmatlcs, 2 g Talk, Text and

. .Ve Da ta.. Methods for Analysin
r

.SIlverman, D . (1993' ) Interpretmg Qual ltatl

Interaction . London: Sag
e. Sue Wilkinson

INTRO D U C I N G FOCUS G R O U PS

The inclusion of a chapter on focus groups in this second edition of Qualitative

,

Research reflects the huge gain in popularity of this method across the social
sciences over the past decade or so. Focus groups are now also thoroughly
familiar to the general public, particularly in the contexts of 'public opinion'
polling (e.g. Elliott, 1998; Kershner, 1998) and consumer affairs (e.g. Aubry,
2003; Knott, 1998). According to the media, politicians hone their images based
on focus group assessments (Brooks, 1998), and market research agencies even
recruit 3-year-olds to focus groups to tap into their purported 'pester power'
(Swain, 2002).
Although the use of focus groups in social science can be traced back as far
as the 1920s,1 prior to the late 1970s their main use was as a market research
tool, and most published studies were in the field of business and marketing
-still an active area of focus group research today (Greenbaum, 1998). In the
1980s, health researchers pioneered the use of focus groups in social action
research, particularly in the field of preventive health education, and the
method continues to be widely used in family planning and HIV/ AIDS educa­
tion, as well as in health research more generally (Carey, 1995; Wilkinson,
1998a) . From the 1990s on, no doubt partly in consequence of the 'turn to
language' in the social sciences (see, for example, Gubrium and Holstein, 1997),
there has been a 'resurgence of interest' (Lunt and Livingstone, 1996: 79) in
focus groups. This has created a substantial literature on the method across
a much broader range of disciplines - including education, communication
and media studies, feminist research, sociology, and social psychology (see
Morgan, 1996; Wilkinson, 1998b for reviews).
Focus group methodology is, at first sight, deceptively simple. It is a way
of collecting qualitative data, which - essentially - involves engaging a small
number of people in an informal group discussion (or discussions), 'focused'
around a particular topic or set of issues. This could be, for example, young
women sharing experiences of dieting, single parents evaluating childcare
facilities, or fitness instructors comparing and contrasting training regimes.
The informal group discussion is usually based on a series of questions
(the focus group 'sched ule'), and the researcher generally acts as a 'moderator'
for the group: posing the questions, keeping the discussion flowing, and

176
 FOCUS G R O U P RESEARCH
S U E WIL KIN SON " INTERVIEWS

to Parti.ci ateiew fully. Although ato focus groups ask are particular population, or simply on the basis of shared characteristics or
enabling grouperrmeedmbtoersas 'gr der
f � s' the mo er, facilitates group r doe s experiences (e.g. middle-aged men, sales assistants, sufferers of diabetes). In
sometimes ref h focus group ou
not
a � i ��
p o : � inetuber
p ;n, but, rath addition to (or instead of) a set of questions, the moderator may present group
questions ofaeactively encouragmg to interact wittial h each other. members with particular stimulus materials (e.g. video clips, advertisements);
discussion, � between researcf �Yti�P�ts s and the poten analytic and in addition to (or instead of) discussing particular questions, they may
This interactlOera n ark ' of foc us group be asked to engage in a specified activity (e.g. a card-sorting task, a rating
__

n - has been Iescn'b ed as the 'ha llm

use of such int an,ctio 1988: 12) . ISC SSlO. n IS' recorded, the data transcribed, exercise). Kitzinger (1990) provides examples of a range of such activities
research. (Morgthe focu s group d' Unal techniques for qualitative data: most in the context of researching AIDS media messages. The moderator may be
Typic ally , . t' 10 . , then, relatively directive, or relatively non-directive.
and then anaIytzed usmg conven . l ·
SIS. ·F cus 0gro ups are . mctlve
dlst .
Focus group proceedings may be audio taped or video taped, with or
tlc ana
t r thema a collY ection (i.e. informal group discussIOn) ,
,

0
commo.n1y contheenme 0
thod f dat aI . It is this perhaps, which leads most without the use of accompanying field notes. Data transcription may be more
for
primanly for the metho d of data an YS1S empha;ize how to run an effective or less detailed ranging from simple orthographic transcription, which
rather than meth d t 0
.

0 . � pleth?ra preserves just the words spoken, to the 'Jeffersonian' form of transcription
contemporarrat y acc,ounts of the w o ana l ze the res ulti ng dat a. The re IS used by conversation analysts (see the Appendix to this volume), which also
focus group, thehermetho than:� 1 an� proced ural choices entailed m settmg preserves a range of linguistic and para-linguistic features, such as restarts,
ce on 0 �
oglca
of advi
con d uc mg
t· a foc us gro up (th e nex t sect ion
. of this chapter provides .
Ices
overlapping talk, pauses, pitch, volume, and intonation. Data management
up and n ne on th the ore tica l
almost � ret�mg focus group data - an emphasis I and epi stem olo gic al chO may be undertaken by hand (i.e. involving cutting and pasting sections
abriefreview) ,bulyzt ing and mterp of transcript) or computer assisted (using programs such as NUD.IST or THE
entaile d in ana ETHNOGRAPH). A wide variety of different types of data analysis may be
will attempt to nsredWh re�s here. ' 11 1o k first at issues related to research undertaken - including content, thematic, ethnographic, phenomenological,
In the sec tio lC� fol lu�'v , I n at greater length - narrative, experiential, biographical, discourse, or conversation analysis
and dat a col lec tIO n usm t� 0 us � o�p s
design ted to data analysIs,? toget�er WI';h some examples drawn from �nd the _

(several of which are discussed in more detail in other chapters of this volume).
at issues rela ch. I will consider some of the issues involved in data analysis in the second part
my own resear of this chapter.
ION
AN D DATA CO LLE CT Practical aspects of conducting focus groups
RE SE AR CH DE SIG N
The focus group literature includes a substantial number of 'handbooks' which
Th e flexibility of foc
us group research offer detailed information and advice about the process of setting up and
rea so for the con tem por ary pop ularity of focus gro ups running focus groups. The most recent of these include: Bloor et al. (2001);
Un dou bte dly , on e r: � exl'bTt 1 1 ° f the me tho d. Focus groups tatibeve
can Krueger and Casey (2000); Morgan (1997); and Morgan and Krueger (1998).
in soc ial scie nce res ear ch IS the Y or co m .bl·ned with quaantireview). I have also written elsewhere about such practical matters (Wilkinson, 2003a,
e qu 1
na tIVe met h d ,0
use d as a sta n d -a 1 on � ro'�ect (see Wilkinorsonin ,the 1999 for 2003b). The handbooks emphasize that obtaining high-quality focus group
techniques as partdofwl�thm rr:ultt�me ��ahod ssr � o /lab orato ry field, to st.udy data depends, at least, upon an effective moderator; and upon a well-prepared
Th ey can be use mp t to c h ang e . t i e in action researchtheprorejectares session.
the soc ial wo rld , or to atte ' us group projectand , Ideally, the moderator should have some basic interviewing skills, some
(Johnson, 199 6).l ChO At . almost ever�:.t;���/s:nf�� se of this flex ibil ity variety knowledge of group dynamics, and some experience in running group dis­
me tho do log ica Ice s to be . ma collection focgaus
s of cussions. Although some of the skills involved in moderating a focus group
can readily be seenforbyexaperu �mg :�: f the recent edited(1999
group research: mp e, B ur and Kitzinger ) and Mor n are similar to those involved in one-to-one interviewing (e.g. establishing
rapport; effective use of prompts and probes; sensitivity to non-verbal cues),
(1993a). proJ.ec� can l�VOo1rvee arna single group of par ticipants meetinged the number of research participants involved in a focus group requires more
A focus group on, or It can
'

mv ny groups' witen h single or repeat in terms of active 'people management'. To enable full participation, the
on a sin gle occ asi
. o1ve as few as two or as many as a doz or ,so, particilstm pan ts moderator may need to encourage quiet participants, to discourage talkative
meetm. gs. It can mv r d e�' ht) Thessecluparbsticiorpan ts may be pre -ex ..
g ones, and to handle any 'interactionally difficult' occasions: these matters are
(the norm is betplewe(e.gen. fou or the
work teams), ative of a y may covered extensively in the handbooks, particularly Krueger and Casey (2000).
groups of peo ether mespemCl�'f�lcars 11ryfa� i1ie
for th� res e�r ch, as represent Because of the number of participants, ensuring confidentiality is a particular
be brought tog
1 79
1 78
 FOCUS G R O U P RESEARCH

S U E WIL KIN SON INTERVIEWS

example, through the consensual pilin g up � f '
fl � e detaIl, - the height of shoe
also ethical issuesmbspe tional nature ofdatfoce, usor
cific to the intetoracsile heels, the size of buttons' the fast� , ng e� ,amsms o� jewelry, the dangers
issue, There aremp le, group me orersevemanythecolres lude nce, intimi 3), of beauty appliances _ focus r � r;:r clpants, w�th multiple sclerosis
groups: for exaticular pant, ear che r (Gr een et a1., 199 h provided a cumulative account �f theP Pun�ane, dally
harass one par r shouldparbeticipre pared to deal approp riat ely with any suc hassles' of living with
the disease (Lyons et a1. 1995 ' 24_5) ' �°wever, focus group participants do
The moderato not always agree: 'they also ml'su nderstand one another'fq�est'Ion one another,
eventualities. handbooks provide guidance on planning and implementing try to persuad e each other of the justice of th ' . ��;�
c

' o · Vl�W and sometimes

Most of the sion itself: the six-volume Focus Group Kit edited by Morgan they vehemently disagree' (Kitzinger, 199:� 1 ), DIsagreements
the focus group ses
larl y com pre hen siv e. Sub stan tial advance prepa­ �nd
ege r (19 98) is par ticu
and Kru pays off in focus group research, To ensure attendance on the e
challenges are also effective in provoking th . deve�opment and elaborahon
ration alwaysessary to over-recruit by about 50% , and to issue reminders, of accounts, For example in a proJ'ect °� understandmgs of HIV / AIDS, which
day, it is nec eshments, all necessary materials (from focus group schedule used focus groups based' on pre-exIs, tmg t
soc'Ia1 ne 'wor1(s, participants. often
The venue, refr and pens), and the recording equipment should be prepared challenged each other on co n tra IC
d' IOns
t' between . h�t they. cl�lmed to
c

to name bad gesThe logistics (particularly arrivals and departures) and the ,
belIeve and how they actually behaved: 'how bout thaJ: hme you dIdn't use
w

and checked, procedure for the session should be run through, 'Ground a glove while taking blood from a pat' t"7'1(: ,. �' at about the other night when
timetable andriefing should be considered, One or more assistants are highly you went off with that boy at the dis��� ( �m�er, 199�b: 1 �5). Challenges
rules' and deb act as recording technician(s), to escort participants to and like these _ in forcing people to defe�d or JustIfy theIr actIOns or beliefs
desirable to and to deal with any unforeseen circumstances, thereby allow­ often generate elaborated accounts , . between members of
from the room, cher to concentrate on effective moderation of the group The relatively free flow of discussion and debate
_

a focus group offers researchers an exce 11ent opport. um y for 1leanng 't ' 'the
ing the resear language and vernacular used b res ond ts' �;1' �; ri� resr:on� ents 'who
discussion, may be �ery different from the�elv�s' (B:�' 19, �; � �, �lstenmg m on focus
group dIScussions or 'structured eavesdr?r:pmg; (Powney, 1988) -promotes
OF FO CU S G RO UP S
OVER I NT ERVIE WS
familiarity with the way research partlClPants habitually talk, and the,
dis tinc t adv ant age s over one-to-onqui e inter­ particular
, idioms ' terminolo gy, an d vocabulary they typIca ' 11y use, ThIS' IS
, contexts
Foc us gro up s hav e a num ber of lect ing data relatively grockly partIcularly the case for focus grouPS cond ucted with'm commumty
ws . Mo st obv iou sly, the y pro vid e a wa y of col ups (see Krueger and Kingd 1998) '' indeed,' SOme focus group researchers have
vie of res ear ch par tici pan ts? Mo re imp ortantly, focus on) argued metho maY be partICUl arly useful m' work WIth ' severely
from
are mo
a larg
re
e
'na
num
tur
ber
alis tic' tha n inte rvi ew s (i.e , clo ser to eve ryday conver - such,
sati , that the hard-to-reach
dIsadvantaged, social roups, people :-vho may be 'uncom­
lude a range ofteacom municative procesllen ses fortable with individual interviews' bufha Y t,o,talk WIth others, particularly
in that they typjokicalingly, inc uing, boasting, upsininteraction, as participange,ts and
g, per sua sion , cha others they already know' 'in the safe an, lf�mlhar context of their" own turf'
as storytelling, The dynarg amic quality of abo gro dis­
(Plaut 1993' 216) " Groups accessed m thIS way inc1ude 'h'Igh-nsk famIlIes ',
disagreement. (sometim ee ut key issue es,'theis feel generally a stri kin g ,et a1. ' city
,m an mner
cuss, debate, andgroups, whes)ichdis, agr at times, may havn misconception of rap ses sion s (Lengua et a1 1992) ' blac. k gay men (Mays et a1., 1992); and
village women in rural counti�s of Chma ' (Wong et a1.,
feature of foc(Jarus
with friends' edrett re is a commo in the context tha
, 1993: 194 ), Theinti t peo ple
group Simply by VIrtue , 1995) ' involved,
' of the number of part'IClp' �nts sImultaneously
will be inhibitfact,abo ut revealing wema te details ng 'sensitiofve'a top ics, focus groups inevitably reduce . the research,er s control over the m' terachon, '
discussion, In contex focus groups arely facllilitasuited to exploridisclosures (Farquhar ma1(mg' focus groups a relativel� 'e :l't�:��� ' �et�od, T�:iS fe�ture has proved
and the group9; Frith,t 200 may actual ing to teKisper sonal mple, it is especially attractive to femini�t es� r� w ave" Istor�cally, expressed
0

with Das, 199ng people to0),talkAccfree ord slin g (1996), for exaup text 0
concern about the power dynamics f the one-to-one mtervlew e,g. Oakley,
easier for you-to-one interview with an ly about menstche ruation in a grority' con 1981 , Finch, 1984),' focus group research h�: �ven sometl" n:es been considered
than in a one to decrease their discomadu lt resear r: the 'solida
fort with the topic,ct to and build upon
am ong akin 'to feminist consciousness raising (I r ert," 1989; Fme, 1992), Reduced
friends seems interactions also 'allow res dents to rea c effect' (Stewart researcher control enables focus group part , Icipan ts to follow their own
Focus group other group members', crepon ergisti agendas and to 'develop the themes most
'
Import an t 0
t them' (Cooper et a1.,
993), One particular benefit of th'I IS, t draw researchers' attention to
the responses ofni, 199 0: 16) , This often leadsg ato'syn atin
the production of more l 0
previously neglected or unnoticed p�enomena, For example, a focus group
and Shamdasaounts than are generated in individual interviews, In the context with former LSD-using" adolescents uncovere d the pos Sl' b l e use o f
' R '
obltussin
elaborated acc and support, one or more focus group members may enthU­ (a strong cough med lcme ) as a substitute , for LSD, In characterizing this
of agreement , elaborate, or embroider an initially sketchy account: for
siastically extend 181
180
 FOCUS GROUP RESEARCH

SUE WIL KINS ON INTERVIEWS

chers typically addressing the issue of 'what is going on' between the participants
.dIscovery as a new piece of the contemporary drug puzzle , the resear in some segment (or segments) of the data, in greater analytic depth and detail.
comment: The two different approaches relate, of course, to different types of research
, we learn question -. and the 'results' produced by the two types of analysis look very
this is one place wh ere f,
0 cus . groups
· . 1'hrough group interaction . ' d
shme
for dru g-expenence different. An initial sense of the distinction between them can be gained from
ed before is a significant issue ic, it is clear that a project on heart attack risk factors, which utilizes both types of analysis
the way the group t a1<es
ethi ng w e ha dn t notic
that som up the top
young p eople. , , , Fro:n , on, something significant to them. A new pIece 0f
.
(Morgan and Spanish, 1985). In this project, a content analysis is used to
gomg
something significant IS and MacD onaId, 199 5 ' 80) address the question of how often different risk factors for heart attacks are
territory is r evea1 e d . (A g ar mentioned, and what these factors are, while an ethnographic analysis is
rch, then rna Y res ult in un exp ect ed ins igh ts: Many of these used to address the question of exactly how (and could also perhaps address
Focus gro up resea arisen 'm. tl1e context of individual intervIeW S. . why) risk factor information is introduced and discussed, in the context of these
are unli. kemov
1y to have
on to cons 'd er '
Is1
, sue s r elating to focu s grou p data ana1YSlS. particular focus groups.
I now e Content analysis is based on examination of the data fo):" recurrent instances
of some kind; these instances are then systematically identified across the
data set, and grouped together by means of a coding system. The researcher
has first to decide on the unit of analysis: this could be the whole group, the
. en siv e ad . e on 11 0W to con du ct
vIc focus group dynamics, the individual participants, or (as is most commonly the case)
com p � red � lth the ext
As indicated, is relatively Itt1e l. the foc. us group literature on hodbo w to �lyze
an the participants' utterances (Carey and Smith, 1994; Morgan, 1995). The unit
groupS, thereg data. Dat a analysl� s sectiOns 0f focus group 'han ly) oks are of analysis provides the basis for developing a coding system, and the codes
the resultin y bne. f, and most commentat ors suggest (or imp are equat11hYe t 11at are then applied systematically across a transcript (or across several transcripts
if more than one focus group has been run). Morgan (1997) proposes three
typically verSUlt. able for analyzm. g one-t0-0ne interview data stud'les,
techniques or use with focu� grou1 data . I n �blished focus grouphow they distinct ways of coding focus group data: noting whether each group discus­
applicable f often omit, or bnefly g oss over, � e det;ils of exactly sion contains a given code; noting whether each participant mentions a
researcherstheir analyses. given code; and noting all mentions of a given code (i.e. across groups or
conducted despite the growmg . b 0dY of literature which considers me thod ­ participants). Once the data have been coded in one (or more) of these ways,
Further, qua1· taflVe method -as theory (this volume,' see also, 00, 200ple,1 '
1
for exam the question of whether to quantify them is a further issue - many qualitative
articularly H01stem,
PGub · 1997'' SeaIe. ettis'a1ts., 200 3'' Silverman, up 1997, 20 f ' 1 t 0
researchers would, of course, argue that the most valuable features of
rium and 0, 2001) s cia l SCI en us .
mg f 0 cu s gro s often aln the qualitative data are thereby lost. Morgan (1993b) argues for the use of simple
200
Wilkinson, method of chOl �
' ce WI'thl' n a clear theoretical frameworork. exp O . 'descriptive counts' of codes, i.e. stopping short of using inferential statistical
locate their. Ons focus group research ers have attempted to justify s 0f qulam tests, whose assumptions are unlikely to be met in focus groups. These counts
rare occaSi of method, they have tended t? do so in relation to normqualitatian-ve are an effective way of providing a summary or overview of the data set as a
their choice OSI. tiVIs
. . rese ar ch , rat he r th an m re lati on to alternativecial sCIe. nt1' f'lC whole.
t
tl'tative or Ps. In partlc. u1 �r m. focus group research, as in so ytic method.s By contrast with content analysis, ethnographic analysis is rarely systematic
approache ore generally, It :IS Stlll. comyarativelY rare for data anal d expliCIt or comprehensive, in the sense of ranging across the full data set; rather, it is
research mn as emb0dying theore tlcal chOl'ces, rarer still to fingoal m' the much more selective and limited in scope. Its main advantage is to permit a
to be see of the pistemologlcal . f d;t:� B/ conmak trast, my retical detailed -more or less interpretive -account of mundane features of the social
discussion CUSS�iOn of data anatY�lCtu:t:ate i i to e such theo world, whether this relates to processes occurring within the focus group
diS
following logical assumptiO. ns expliCIt. . . itself, or whether (as is more typical) talk within the focus group is seen as a
and epistemo means of access to participants' lives. Ethnographic analysis aims to be
contextual, i.e. to ground interpretation in the particularities of the situation
'ethnograph ic' analy
siS under study, and it aims to represent the social world from the participants'
'Con tent ' versus
t approaches to perspective, i.e. to ground interpretation in participants' views of 'what is
d' .
Ist '
mc f on be tw een
1 tw o d 'ff
I ere n
ake
I want to fomcuS a key
p data: c t 1y sis and ethnogra
phic ana Y �v.e
l SIS
. 3
going on', rather than the analyst's view. Data are generally presented as
grou
analyzinganal SIS. produces a �;:�ti:��y systematic and comcorp prehens accounts of social phenomena or social practices, substantiated by means of
Content orYoverVI. ew 0f the data set as a whole, sometimes in e selora t.mg extensive illustrative quotation from the focus group discussion. Key issues
summary ve element, whereas ethnographic analysis is mor ective, in ethnographic analysis are how to select the material to present; how to give
a quantitati
_

183

1 82
SUE WILKINSON INTERVIEWS

A
FOC US G R O U P RES E RCH
due weight to the specific context within which the material was generated, of women's talk ut the causes of health �nd '
while retaining at least some sense of the group discussion as a whole; and
how best to prioritize participants' orientations in presenting an interpretive a.nalysi� systematiabo
cal ly rec ord s the of .
me
ll!nes;. .r-r�wever, the first
ntI
5
.

account. (mcludmg null categories) summan. zm. g what ons wIthm eac, h category
nu mb er
t he second records the words'in wh o h these 'mentI. ons
IC
these" 'ment'IOns are, while
A particular challenge is substantively to address the interactive nature of them as quotations under each category head'mg (ex ' are �ouc1led, presenting
focus group data: a surprising limitation of focus group research is the rarity cludmg null categories).
with which group interactions are analyzed or reported (Kitzinger, 1994b,
Wilkinson, 1999). Extracts from focus group data are typically presented as if Bo x 1 0. 1 :
they were one-to-one interview data, often with no evidence of more than
one research participant being present;4 still more rarely does interaction per \"<'U�ie5 of breast

se constitute the analytic focus. This is all the more surprising given that -as (1) �TITATIVE VE RS ION
noted above focus group researchers typically emphasize interaction 1. Infection : 0 instan ces
between participants as the most distinctive feature of the method, even 2. Hereditary or familial ten
2 instances
cautioning that researchers 'who do not attend to the impact of the group family history (x2)
dencies:

setting will incompletely or inappropriately analyze their data' (Carey and 3 . Agents in the environ

�
ment:
Smith, 1994: 125). �
(a) ' OiS? S ', working condit
fb I
ion, climate: 3 ins
l n l um pans; exposure to sun; chem ic tances
a um
als in food
"""
)' / drugs or the
con traceptive pi":
Unpacking content analysis 11<>"' ''''''''' and taking the con traceptive pill 1 instance
variants)
4. Secondary to other diseas
es: 0 inst anc es
The majority of published focus group studies use some type of content 5. Stress, strain, and wo
rry: 0 ins tan ces
analysis. However, given the wide variety of terminology employed in these 6. Caused by childbear
ing, menopaus
not b reast feedin g; late chi ld b eare: 22 ins tan ces
studies, and their frequent - sometimes extensive - use of illustrative sin glelnot hav ing chi ldre n' hor in g (x3)'' h aVlng . onl y one chi ld; b ein g
quotation, it may not readily be apparent that this is so. To appreciate the m
��
un specified (x4); flattened ni pp 1 a�I ' tr?� b le W ith ?reast feed ing -
underlying theoretical similarities between most analyses of focus group discharge (X2) ( 2) , Inverted nip ple s (x 7);
nip ple
data, it is necessary to appreciate that content analysis need not employ a 7. Secondary to trauma or
formal coding scheme, nor need it be a precursor to any kind of quantification. to surgery: 9 instances
knocks (x4) ' unspec'f ed InJu
At its most basic, content analysis simply entails inspection of the data 8. Neglect the co;stra
' . r�; ai r getting ins ide body
ints �� poverty. 0 Ins (x4)
for recurrent instances of some kind, irrespective of the type of instance (e.g. 9 . Inhere nt suscep tJbllt
1 O. Beha �lOr, own responsib
. " ty, . tances
. tances
word, phrase, some larger unit of 'meaning'); the preferred label for such
individual and not heredi
tary. 0 Ins
ility: 1 instan ce
instances (e.g. 'items', 'themes', 'discourses'); whether the instances are sub­ miX ing specific foods
1 1 . Ageing, �atural degenerat
sequently grouped into larger units, also variously labeled (e.g. 'categories', 12. Other: 5 Instances ion : 0 instances
'organizing themes', 'interpretive repertoires'); and whether the instances 'several thin gs' ; 'a Jot'; 'mu lti-f
-- or larger units - are counted or not. In the sense, then, that most analyses of cancer; 'an yth ing ' cou ld wa k actona . l ' ,. everyb ody has a 'do rma
nt'
focus group data report recurrent instances of some kind, and do so more e a dormant can cer
or less systematically, they are essentially content analyses. (2) QUALITATIVE VE RS ION
To illustrate this point, I show in Box 10.1 two different content analyses­
one quantitative, one qualitative - of the same piece of focus group data. The Hereditary or familial ten
dencies
'results' of the quantitative content analysis are presented as frequency counts, 'I me an the re's no fam ily histo.QI
'
while the 'results' of the qualitative content analysis are presented as Agents in the environm
illustrative quotations. The data are drawn from a segment of a focus group
ent:

in which three women who share a breast cancer diagnosis are talking about
(a) 'poisons ', working conditi
on, clim ate

possible causes of the disease (see Wilkinson (2000) for more information about 'I was once told that if you use
'Lo oki ng years and years a the m alu mi n .l um pa ns that cau se cancer'
0 I
� �:�d �
this focus group, and the larger project of which it is a part). Both analyses verYb odY use d to sit ab out
take the 'mention' of a cause as the unit of analysis, and organize these sun nin g themselves on the
cancer from sun shi ne' �ac O W all of a sud de n you get
'mentions' using a category scheme derived from Blaxter's (1983) classic study
184

185
1m iii

FOCUS GROUP RESEARCH

SUE WIL KIN SON " INTERVIEWS

2003). We may also note that neither of these analyses has preserved the
interactive quality of the focus group data, although it is possible for a thematic
' I don 't know (about) all the
che mic als in wh at you 're, eat
ing and thin gs these
analysis to do so (see, for example, Braun and Wilkinson, forthcoming; Ellis,
ivated and everyth ing 2002).
days as wel l , and how cult
eptive pill
(b) drugs or the con trac
'
k the pHI at a you nge r age
'Yo u kno w, obv ious ly I too
, menopause
Caused by childbearing
� l S one thin g that you cou ld be wary of
'
,
So far, we have seen that content analyses which look very different (i.e.
' Inverted nip ple s, the y �ay t
�
at h a t
0 ac ua li trYi ng to breast
feed I did n't rea lise I providing quantitative or qualitative 'results') in fact treat the data in the same
' U ntH I cam e to the pOi nt
/ the m was nea rly inve rted or whatever, kind of way (i.e. inspecting them systematically for recurrent instances). We
�� �
d one
had flattened nip ple s an it and I wa s several weeks
can further note that these various types of content analysis share a similar
�� :��
a nd
h
b r St d'
so I had a lot of trOU bl�
underlying epistemology: one in which research particip.ants' talk is taken as
on my
ng ge i t so t at he cou ld suc kle
wit h a breast pump tryl
nip ple , I did have that pro ;
ble n '
'
r anythin g like providing a 'means of access' to something that lies behind or beyond it. In
' Over the years, .every,
I
� � \� :!
cO U ld sa ' t h a ene d mo nth ly o
a dressing my content analyses of women's talk about the causes of breast cancer, the
J
tha t, it WOU l.d U t s
��� �� �:;� �
th b iS I
�
ag e nothing to put
som ewh ere words of focus group participants are taken to provide a 'transparent' window
on or anyt h In 9 I�lke t, tt it �as , it wa com ing from
, onto what they understand, think, or believe about, say, the role of repro­
ductive factors in the etiology of breast cancer? Similarly, in the AIDS Media
:; � � ��:
tly cru s
and it were just kind of gen es a
hiC h you have chil dre n mak
'I me an, I don't know W heth he g
-
h a my eight year old rela
- tively late, was
I Research Project (Kitzinger, 1994b), mentioned earlier, and designed to
difference as weII b eca
u
investigate 'audience understandings' of HIV / AIDS, participants' words are
an old m um
: one that you 're more likely
to get it than if taken as 'revealing' their understandings or beliefs about, say, the mechanisms
of HIV transmission. In such studies, self-report is used to infer the relatively
had -
'They say th at If you 've only
you have a blg fam ily'
stable 'cognitions' (beliefs, attitudes, or opinions) assumed to underlie people's
Secondary to trauma or
to surgery
talk (and at least sometimes -to inform their subsequent behavior), to which
'So me tim es I 've heard th �
t lsnocks can
�;�:� :����
o
the researcher has no independent access.
?'
iS . . . you kno w these
'I th e n rem em ber ed tha t I d ban g� d my t
A similar epistemological status is also commonly afforded to talk in focus
.
sho ppi ng bag s with a woo
den rod thin g, those big trol
ley bags . .
loratory, It may group studies which are designedly ethnographic (rather than content analytic)
�\� �� � ;� � in nature, i.e. studies which aim to provide contextual, interpretive accounts
exp
�
'I to hos pita l eve n for an
'I always think th e I
se ems to me that it's
be all wro ng, u t i
:����: ;� ���:�r:�
'
ell the. air ets
-;- find
to it, it
tha t the re's mor e to it than of their participants' social worlds. For example, in Lyons et a1.'s (1995)
n ot 10 ng afterw
fO t
der if the air getting to your
inside study of women with multiple sclerosis, and Agar and MacDonald's (1995)
they th oug ht , y . ' d , study of ex-users of LSD (both referred to earlier), research participants' talk
. . . brin gs on can cer In
any form
is taken to provide a 'transparent' window onto the circumstances of their
Behavior, own responsib
ility
the sam e me aI '
lives outside the focus group (to which the focus group moderator has
' I wa s also told that if you
eat tomatoes and plu ms at no independent access), and which are inferred from self-report. What people
say in the context of the focus group discussion is seen as 'revealing', for
Other
a can cer , il d lL it's a example, the nature of daily life for people with chronic physical illness, or
'H e told the m nurses in his
lectures that everybody has as flagging up a 'significant issue' in the life 'territory' of the drug-experienced
t'
young. In other words, talk is used as a 'means of access' to something that
���� �; ���:� � �a:� :�:��
d
actoria l'
'I d t n o l n e, can it? It mu st be mu lti-f
lies behind or beyond it, rather than treated as of interest in its own right.
I want to contrast this view of talk (i.e. talk as a means of accessing a pre­
given social - or psychological - world) with an alternative one: one which
h both are
ferenti althougseco considers talk as constituting the social world on a moment-by-moment basis.
These two content eanal yses, .the�h I���I �:rrfradifmew or (. The nd type, This is to give talk a very different epistemological status from the one presented
m the sam un � erlY mg e t a so far - and one which radically effects both the kind of study undertaken
derived. fro l'1tatlVe
reportlng qua. ' data,1ysl 0
'thematic' analysquisota(sotlO.me-nS
IS , ften d escn'bebde aspreasen and the kind of 'results' obtained. Affording such a status to talk opens up
times as a 'dISCo urse' ana ' s)6h'anandm. tma y ted. wl'th the lkinson, the possibility of seeing the focus group discussion as a social context in its
integrated into the text, rather t abular form (e.g . Braun and Wi
187

186
SUE WILKINSON INTERVIEWS FOCUS G R O U P RESEARCH

and, further, the possibility of subje�ting it .to direct observat�on (Puchta and P.atter, 1999); displ ying opinions (Myers, 1998). While it may be
own right,
(rather than studying it in order to infer more dIstal. soclal - or psycho :oglcal useful t? c?nsider what goes on�In focus group qua focus group, analysis need
phenomena) .8 The resulting study will nece sanly be ethno � rap � Ic, �nd not be lImIted by the specificity of this particular context: it can also address
�
�
socIal proce s es. W nle thIS fm .ght more generic conversational phenomena, ranging across the practices and
_

will provide a detailed, contextual accoun t of �.

seem an obvious epistemological move to anyon e famIlIa r wIth theore tical actions displayed in the talk (Schegloff, 1997).
work on language, or with contemporary social scientific de�ates about esse.n­
Holste In,
To illu�trate this point, I offer below a sample data analysis which I would
charactenze as broadly ethnographic, and also as based upon the principles
tialism and social constructionism (see, for examp le, Gubnu m and
1997; Kitzinger, 1995), it is a radical propos : al fo most f cus group researc hers, of ethnomethodology and conversation analysis. That is, it seeks to offer
.� (most) a detailed interpretive account of 'what is going on' within the talk which
including those working within an ethnograph� c tradltIon. Even though
ethnography is predicated upon direct observ atlon, few focus gro � p resea �chers constitutes the focus group; and it theorizes this talk as action oriented as in
conducting (broadly) ethnographic analys is have tur r: ed theIr attentI On �o pursuit of particular, local interactional goals. '
observation of 'what is going on' in the focus group ltself; and fewer �tlll The data on which this analysis is based are presented in Boxes 10.2 and
have paid detailed attention to talk as constit utive of social or psycho logIcal 10.3. These two extracts are drawn from the same segl1lent of the same focus
- life. gro�p as used in the content analyses above, i.e. the part of the discussion in
In exploring this kind of approach, I have found the theoretical fra�ework whIch the three women are talking about possible causes of breast cancer.
offered by ethnomethodology and conversation analysis to. be partIcu .larly Here, I have identified myself ('SW') as moderator of the focus group, and
valuable (see Chapter 12, this volum e, for a more detaIle d o .vervIe ,:,,). I have called the participants 'Freda', 'Doreen' and 'Gertie' . Note that the
Conversation analysis assumes that it is fundam entally through mterac tlOn lev�l of detail p�esented in the transcript is appropriate to the level of analysis
that participants build social context - indeed, it is often explicitl� referre d WhICh follows, l .e . It. IS. somewhere between a simple orthographic rendition
to as the study of 'talk-in-intera ction'. It therefo re seems extraor dInary th �t and a full 'Jeffersonian' (conversation analytic) transcription.
focus group researchers looking for a way to analyz e the key fe � ture of t�eH
. d Data extract 1 (Box 10.2) opens with my question (as moderator) about
data, i.e. interaction between participants, have not more extensIvely an utlhze . causes, and the responses from Freda and Gertie to this question. Note that a
this approach. Central to an ethnomet�od�logical or . conversationpartIcu �lyhc content analysis (of the kind presented earlier) might code Freda's initial
framework is the notion of talk as actlOn, l.e. as deslgned to do we can lar respon� e (lin� 3) as 'I don't know', and Gertie's subsequent response (lines
things within a particular interac tional contex t. Within a focus group 5-9) as Items In the categories 'agents in the environment' (aluminium pans)
see how people (for example) tell stories, joke, agree, debate, argue,t chal�en ge, and 'behavior, own responsibility' (choosing to eat tomatoes and plums at
or attempt to persuade. We can see the ways in which the � presen partlcu lar the same meal). A discursive-ethnographic analysis, by contrast, focuses on
'versions' of themselves (and others) for particu lar interac tIOnal purpos es: (for t�e immed�ate interactional context. Talk about causes can be interactionally
example) to impress, flatter, tease, ridicule, complain, castigate, or �Itself, ondone. �ncky,.partlcularl'y when a presumed 'expert' is asking questions, or in settings
Participants build the contex t of their �alk in a
. r: .d t hrough �hat talk on m whlch potentIally equally knowledgeable others might have differen
a moment-by-moment basis. The talk Itself, In ItS Interac tIOnal contex t, ca � or even conflicting opinions. Conversation analysts (e.g. Sacks, 1992: 340-7)t
provide the primary data for analys is. Furthe r, it is possib le to harnes s anal � hc ha:e noted the asymmetry between being the first to express an opinion and
resources intrinsic to that data: by focusing on participants' own understan .dmg �eIng second - i� that going first means you have to put your opinion on the
of the interaction - as display ed directly in their talk, throug h the conver sat :� nal lme, whereas gOIng second offers an opportunity either for agreement or for
approa c� . zes
pnont potent�. al challe�ge. Consequently, speakers often try to avoid first position,
practices they use. In this way, a conver sation analyti c �
the participants' (rather than the researc.her's). analysis of the :nteract IOn: and thIS IS. preCIsely what Freda does in response to the moderator's ques­
In all ethnog raphlc analyse s). . she declines to give
hon:
a broadly ethnographic goal (if not one achIev ed an opinion, and bounces the question right back
The traditions of ethnom ethodo logy and conver sation analysi s genera lly to the moderator, as a 'counter' (Schegloff, 1995: 7-10). It is not simply then,
rely primarily upon the use of naturally occurring data, i.e. data produc ed as a cOntent �nalysis within an essentialist framework might suggest
independently of the researcher. These data, however, enco�passgenes) a. ran�e F�eda ,�oesn t know' what causes breast cancer: from the perspective, that of a
of institutional settings (e.g. classrooms, courtrooms, doctors sur'ever day' , in dlscurslve-ethnographic analysis within a social constructionist framework
which talk has been shown both to follow the conventions of H nta � �he is not here reporting a state of mind, but is engaged in a piece of local
conversation and system atically to depart from these (Drew and � �e, mteractional business.
SOCIa l The moderator (SW) avoids answering Freda's direct question: instead
1992). Likewise, data from focus groups range across both 'everyday t b she reformulates it (in the manner typically recommended for intervie
actions (e.g. arguing, joking, teasing, complaining) �nd actions likely questIO
? e
nS and focus group moderators), making clear she is interested in whatwers
specific to the particu lar (resear ch) contex t, e.g. askIng elabor ate the
1 88 189
 FOCUS GROUP RESEARCH
SUE WIL KINS ON INTERVIEWS

ledge. Within a social constructionist framework, however, this attribution

of ideas about cause to folk knowledge is seen as an interactional device seek­
Box 1 0.2 Data extract 1 ing to protect the speaker from challenge (although, here, it fails to avert
1+2)
ridicule).
ext rac t s
SW : BC P 1 2 ( C au s e s Gertie's candidate causes, then, are presented as 'stories'; however, only
a wha t c a� s ed you
r breast moments later, even these 'stories' are retracted. By the end of Doreen and
D ' yo u hav e any i de
01 SW : Gertie's subsequent exchange (at line 32), Gertie, like Freda before her, is
se b r e a s t· can c e r do
o f you .
can c e r [ p au s e ] any
02
N o - What doe s cau
you t h i nk . claiming to have 'no views' on the causes of breast cancer. Again (within this
03
04
Fre :
SW : What do you thi nk
i t migh t� be ? framework), this not simply a straightforward report of a cognitive state:
0 5 Ger : [ cu t s i n ] The re ' s a lot o f go ing it arises out of the interactional sequence within which it is embedded, in
06 abo ut . = I was onc e
told tha t i f you use the m
the course of which both Doreen and the moderator have implied, through
. hh I was a l s o
07 a lum inium pan s tha
t c aus e can c e r .
their laughter, that Gertie's candidate causes are rather implausible; indeed,
e s and p lum s
i f you e a t t om a t o the moderator's probe (lines 19-20) can be heard as 'gositioning' (Wilkinson
08 t o l d tha t i f you -

09 at the sam e mea l

tha t - and Kitzinger, 2003) Gertie as the sort of gullible person who believes any­
1 0 Dor : [ l aughs ]
l the s e tho s e
thing she is told. Gertie responds first by reminding everyone that she is
Hav e you h ear d a l
11 Ger : [ t o Dor een ] not reporting her own views, but those of others, and then she flatly refuses
12 thi ng s ? to offer further candidate answers, explicitly handing the floor to Doreen at
[ l aughs ] No
13 Dor :
hea rd and -
line 32.
Now tha t ' s wha t I
MIn
1 4 Ger : When Gertie re-enters the conversation (at line 65,9 in data extract 2, Box
s eve ra 1
Dor : [ l aug hs ]
15
t h l' n g s tha t if you l i s t en t o 10.3), with a subsequent suggestion of a candidate cause (the theory that cancer
we : : 1 1 -
1 6 Ger : O h the re ' s
people [ pau s e ] is 'dormant' until woken), she is still attending to the danger of being laughed
MIn
17
18 Dor :
at. However, here she deals with the risk of ridicule by using a very different
the y t o l d
19 SW : ' e,
[ t o Ger t l l aug hl' ng l y ] What e l s e have kind of footing10 (Goffman, 1981): the 'dormant cancer' theory is painstakingly
20 you? constructed as someone else's opinion - that of a specified medical expert, a
2 1 Ger : Par don ?
Wha t e l s e hav e the y t o l. d
Dr Patterson, at Springfield General Hospital (the hospital where most of these
l aughin g l y ]
22 SW : [ t o Ger t i e , women will have received treatment). She carefully monitors the reception
23 you ? of this theory, and even though Freda and Doreen affiliate with this view
2 4 D / SW : [ l aught e r ]
thi nk o f f hand I knew a - I knew a lot (at lines 73 and 74), she checks to be sure she has their support (line 75),
25 Ger : I c an ' t
r the yea rs f r om peop l e and continues repeatedly to stress that this theory comes from her sister's
tha t I ' ve hea rd ove
26
who ' ve p a s s e d o n ' Oh yea h we l l tha t causes nursing training: 'she told us that, and that came in her lectures' (lines 77-78);
27
can c e r ' .
'according to him' (line 78); 'that's what she was told' (line 84). The effect of
MIn
28
29 Dor :
all this footing is to emphasize that these ideas are not her own, and that she
3 0 Ger : But I don ' t know but - is not to be held accountable for believing them. Again, within a social con­
3 1 Dor : [ cu t s i n ] I mean
uhm-
have you ?
structionist framework, the attribution of views to others does not offer a
[ To D ore en ]
32 Ger : Now I ' ve no v '
l ews on th i s 'transparent' window on what Gertie 'believes', nor does it indicate the
'source' of her information. Gertie is not simply repeating what her sister
may or may not have told her Dr Patterson had said; rather, she is using footing
particip, ants themselves 'think it mightstbe' (line 4), rath�r tha� in any purpo� �! as a conversational resource, in order to manage the delicate interactional
business of presenting an opinion without sounding ignorant or stupid.
It is wlth thIS reassurance .
:eby putt���
, (i e scientific) causes of brea cancer.labe Doreen then rejoins the conversation (at line 91) to offer another candidate
�� �i off�;s some 'stories
:� ' (Le. folk wisdom� . led as such)� the
pOSItlOn, and attractm Just the k ' d �'
g cause - a 'bang' on her breast. Her story, elaborated in lines 91-113, is
herself in the vulnerable first speakingFred s counte� �nabled her to avOl apparently 'touched off' by Gertie's mention of 'a knock or whatever in the
of second s eaker disagreement thatp, lauga'hs at Gertle s res-po�se. t a appropriate place' causing cancer to develop (line 8n followed by Freda's
Doreen the�hird member of the grou tie's references to 'stones , an �o at � �� subsequent acknowledgment of the theory that 'knocks can bring one on' (lines
within � n esse ntia list fram ewo rk, Ger
ta�en . as transfarent re�o�ts f the­
87-88). Note that just before Doreen begins her story, Gertie and Freda have
she has 'heard over the years', wouldasb�md
source of her ideas about cause, i.e.
Icatmg a rehance on 0 k :now both placed considerable distance between themselves and the 'knock' theory:
1 91
1 90
 FOCUS GROUP RESEARCH

SUE WIL KIN SON INTERVIEWS --- - ---

--
-
--

108 I ' d j us t be t t e r men t i on it' i n c a s e i t t u rned out

2 109 t o be . hh you know s o r t of they ' l l say- , they then

Bo x 1 0.3 Data extract

(-,x tr a c t s Q'5�
110 come round t o me a f t erwards and say ' Ar e you sure
111 you haven ' t d - done s ome inj ury t o yours e l f '
SW RC: P 1 2 ( Ca u s e s
112 >or that sort o f t h i ng< ' c o s you know , it j ust
y Q�c k in the
a nur s e [ pau s e ] wa : 113 sprung t o wing . . hh ' Cos I - I ' d men t i oned it to
1
s 1, s t e r ---
was
6 5 Ger : My .. - she was a t w . hat i s
i s t in c t ] . And she 1 14 the GP and she ' d s o r t of said ' No no . hh it ' s
66 1 9 2 0 s she [ ' nd - i t ra inin g the re
d he r
era l noW . - She d .
. _

115 nearly a lways hormona l ' so i t ' d gone out o f my

spr i ngfi e l d Gen . hh
r s on at the t lme
67
_ ' .
a d oc t or P a t t e
i t ' l l be b e a S;;� '
116 head and a n - bu t - but then she was s aying ' No

who u s e d t 0
and th ere was o l d
68 he t

0
nur s e c 0 . hh and
l e c tur e t o the
.
117 >wha tever< and when i t wasn ' t
69 ry b dy has a
t eve
nur s e s ln, h i s l ec tur es tha 118 a cyst then I s t a r t e d t o think of another c a u se
70 the m -- . wh e t h e r
pau s ' t ' s a case 0 f
e ] 1.

( G ) : MID
[pau s e ] g,
. nd [ 119 you s e e bu t - . hh
71 can c er
an t 120
72 i t l ay s do rm ,

I ' ve hea rd tha t . 1 2 1 Dor : uh : m I - I mean I sup - i f -- i f they knew what l:he cause

7 4 D or : MID
7 3 Fre : Yes
122 was they wou l d - they wou l d b e ab l e t o t r e a t it
t? 123 wou l dn ' t they .
you hea rd tha
7 5 G er : Have
( F ) : MID � 1 2 4 G e r : We l l you know I
76 tha t came in er
l d us tha..1. and
, t hat she t o 1 2 5 Dor : [ cuts in] I don ' t think it c o u l d be or� cause can
7 7 Ger : We l l ye" ng
c

to him any thl

s e ] a c c ord i ng
( G ) : MID
[ pau 126 it? I t. mus t be mu l t i
78 l e c tu r e s . hh and .
i t up 127
c o u l d wake
Do r : MID
79
1 2 8 Dor : . hh mul t i - f a c t o r i a l
80 a t e p l a c e . hh
n the app rop r i
or wh ate ver i 1 2 9 Ger : [ cu t s in] You ' ve heard them s a y -
8 1 Ger : a kn ock tha t ' s wha t -
l d dev e l op but 1 3 0 Dor : whatever t h e w o r d i s
and the n i t wou
Dor : MID
82
83
she was told .
tha t ' s what

8 5 Dor : MID
8 4 Ger :

when 1 -
, Freda saying it is a theory she has 'sometimes heard', but that she has never
8 6 Ger : Bu t .I , ve hea rd tha t kno cks can brl
ng

8 7 F re : [ cu t s
in] S ome t ,
lme s
kno ck s ) [ i ndi s t inc t] had any knocks herself (lines 84-5), and Gertie attributing the theory to her
nev er ( ha d any
88 one o n bu t I ' ve
.
sister's nursing training, some seventy years earlier (lines 65-8 and 78-81).
8 9 Ger : No
don ' t thi nk t ha t ) [ ind i s t inc t ] In telling a story about her own knock, then, Doreen can be seen to attend to
[ cu t s ln ] W e 1 1
' ] (I
9 0 F r e : [ cu t s
In
I ' d hea rd tha t
f r om som ebo dy
else the risk of aligning herself with a belief in knocks, and thereby possibly attract­
1 - when obv i o u s l y
9 1 Dor : ing scorn or censure (see Potter (1996: 142-7) for a detailed discussion of
_
,
th l S was sus

C 1 0US n- I . . the n rememb ere d

u SP1
an d s o when
92
my lump wa s s
" I the distancing, neutrality, and alignment).
93 s . hh uhm [ t c h ] you Doreen never actually says directly that she believes her breast cancer to
br eas t Wl' th thi
I ' d banged my '
lng
94 woo den rod th
95 kn ow the s e s h
op pln ' g bag s w i th a have been caused by a knock to her breast. She simply 'remembered' (line 90)
. hh tho s e big
t ro l l ey b ag s ? having banged her breast, and reports feeling it necessary to 'mention it' (line
F r e : MID
96
101) to her surgeon - thereby further displaying to her co-conversationalists
1 - 1 - 1 >do n ' t a s k
<
97 s e s tup i d thi ngs
me how I do the that she is a rational person who informed a medical professional of the knock
9 8 Dor : r as I was
wee n the car doo
99 bu t I it wed ged bet
got
car I g o t i t wed
ged in the car
in order to check out all possibilities (she has already shown herself to be
100
,
g e t t lng ou t o f the - -
rop i a t e ly S l, z
r ed aware of a range of other possibilities, in her previous discussion of reproduc­
S �
thi s [ p au s e ] app
101 d o o r so i t - so
s i z e of t h l
' l n m n you know wen t tive factors, not reproduced here). The surgeon's reported response, 'You're
d 1- an d I que r
ro d tha t was the
102
br eas t a n
i ed tha t . hh an
d always looking for a reason' (line 99), is a generalized formulation that does
103 int o my kn ow ' You ' re
not dismiss the 'knock' theory specifically, but that even-handedly dismisses
,
s a l' d you
tan t sur g e on ]
f
Mr F e l l [ c ons ul
104 ' [ l aug hs ] d ' you
k now
any theory (actually or potentially) offered by Doreen - and, by implication,
��
o r e a s on
a l ways l o okl ng .
�
mig ht be
105
o fin d s ome thi ng tha t
anyone else. Ventriloquizing the surgeon in this way enables Doreen to present
the 'knock' theory as no more or less plausible than any other theory (to which
o
106 ' You ' ve alw ay � Wel l
u t I thought '
se o f l t you know . hh b
1 07 the cau
the 'always looking for a reason' dismissal is equally applicable).
193

1 92
SUE WILKINSON INTERVIEWS
FOC US GRO UP RES EAR CH

The surgeon's response, with its implicit suggestion that looking for relatively underdeveloped
'reasons' and 'causes' is futile, also provides evidence for Doreen's later claim participants. I have argued herfor the stud f l.:teraCtl�ns between
e that ethn��1 et odologlcal and �on research
.

that 'they' (doctors) do not know the causes of breast cancer (lines 121-122). analytic approaches may prove
If they do not, and if looking for causes is pointless, then the 'knock' theory endeavor. p articul'ar y useful for pursumgversation
such an
is as plausible as any other. Likewise, the CPs dismissal of Doreen's theory is
also reported in such a way that the 'knock' theory is left open as a possible
cause: the competing cause offered by the CP Cit's nearly always hormonal',
lines 1 14-1 15) is explained as having been offered as a cause for a cyst, not
cancer. The reported misdiagnosis has the added benefit of pointing to the Although psychologists Emory Bogardus and W
to develop survey instruments m. thalte
1.
fallibility of the medical profession (re-emphasized in lines 1 21-122). Within techniques r fhurstone. used focus group
.
c

a social constructionist framework, the concern is not with whether Doreen groups IS more often credited to sociolo ist Rob e 1 920 ' , 0f focus
5e , the. ,l �vent"lOn
'really' believes the 'knock' theory, or with whether medical professionals used them to study aud e rt M erto n and hIS
ience reaction t� ra d'10 �rograms in the 1940s. gues, who col lea
2. However, it can be difficu ' consummg to r�crU1t' and bring
appropriate participants. lt and t1me
'actually' dismissed her possible explanation. Rather, it is with how Doreen
designs her talk to illustrate: (1) her own rationality, both in reporting the knock together
3. Morgan (1988: 64) also uses this
'

and in assessing its merits and demerits as a theory; (2) her own willingness narrowly defined, mapping (broterm inolog� but h'�s two types of analysis are more
to listen to the opinions of expert others; (3) the fallibility of the medical divide (Bryman, 1988), adl y ) ont0 t e tradItIo. nal 'quantitative--qualitative'
profession; and (4) the plausibility of a knock as a cause for breast cancer. 4. It is not uncommon to use
study, and one-to-one interviews and focus groups in the
Doreen's final statement - that the causes of breast cancer must be 'multi­ to treat the data derive eacI method as co. mmensurate, witsam e
factorial' (line 128) - enables her to maintain the possibility that her injury · USS10n of the relationship betweed nfrom
dlSC . h no
quoted extracts are derived from whichthem -and often with nom. d'1cah.on ofwhich
1 '

was causally implicated, while not denying the potential relevance of other
5. With the addition
sou rce (e.g. Espm, . 1995; Press,
(more medically approved) causes. In this interaction, then, Doreen designs of an 'Ot he r ' ca tego 199 1) .
6 . See, for example, Weaver and ry.
her talk to display to her co-conversationalists that she is a rational and open­
confusing, given that the epistem Ussher. (1998) . ThIS' termm. ology is potent
minded person.
discursive psychology are more akiolo ���I �s�umph..on� of discourse analysiallis!y
presented in the foll ing section than t ��� n e a ternahve �thnographic approach
ethnographic analysiow s (see also Potter, thi� t:;) ana lYSIS, or even 'traditional'
IN CONCLUSION
7. And elsewhere (Wi .
talk about the causeslkinofson
br �
2000 200 1) I have de;no strated
ast c�ncer t0 women s blO � grap�ies. analyses which relate
In this chapter, I have outlined what is involved in conducting and - especially 8. Viewed in this way, it is not
- in analyzing a focus group, with particular reference to the theoretical in what way talk may be reflevee nf:�� necessary to address the ISsue of Whether and
�: ;:rres�nt�tive ?f something beyond itself.
9. I have omitted a num
and epistemological issues entailed in different types of analysis. Although I
am a keen advocate of focus groups, I would not want to claim that they ber of line �
the discussion Doreen offers aturan�e 0f rep eres s o savmg space. In this portion of
offer 'a method for all seasons' (Tuck, 1976). Focus groups are a method of breast cancer. She ends her rn Y �uggesro�uctive factors as pos sible causes
may be a cause append in tmg that not haVing children at all
of choice when the objective of the research is primarily to study talk, either the eS h
there?'. It is with this encou�age��:t�0 present add on that 'the�e' s more to it than that, isn't
conceptualized as a 'window' on participants' lives or their underlying beliefs Itional causes that Gertie comes
and opinions, or as constituting a social context in its own right, amenable to in again.
10. The term 'footing
' rs to the ran e of relationShI.pS between
direct observation. If, by contrast, the purpose of the research is to categorize they say. It enablesrefe
distinctions to b� m d b tween people makin speakers and what
or compare types of individuals or social groups, in terms of the lives they
lead or the views they hold, then focus groups are less appropriate (although Own behalf, or reporting the claims of �tl�er:. g claims on their
they are not uncommonly used in this way).
Focus group data readily lend themselves to analysis by content analytic
or ethnographic methods (both of which I have illustrated here), as well as by
other qualitative techniques; and the resulting analyses can be presented in
a variety of ways, ranging from numerical tables summarizing a whole data
set, through prose accounts containing lengthy illustrative quotations, to
detailed interpretive accounts of a relatively circumscribed single data extract .
Focus groups offer particular opportunities - although this potential is as yet
1 94
1 95
 FOCUS GROUP RESEARCH

SUE WILKINSO N INTERVIEWS

Bryman, A. (1988) Quality and Quantity in Social Research. London: Unwin Hyman.
Carey, M.A. (ed.) (1995) Special issue: Issues and applications of focus groups.
Qualitative Health Research, 5 (4).
Carey, M.A. and Smith, M.W. (1 994) Capturing the group effect in focus groups:
A special concern in analysis. Qualitative Health Research, 4 (1): 123--7.
Cooper, P., Diamond, I., and High, S. (1993) Choosing and using contraceptives:
Integrating qualitative and quantitative methods in family planning. Journal of the
Market Research Society, 35 (4): 325-39.
Drew, P. and Heritage, J. (1992) Analyzing talk at work: An introduction. In P. Drew
and J. Heritage (eds), Talk at Work: Interaction in Institutional Settings . Cambridge:
Cambridge University Press.
Elliott, L. (1998) Focus groups tested tax plans. Guardian, 24 March: 3.
Ellis, S. (2002) Doing being liberal: Implicit prejudice in focus group talk about lesbian
and gay human rights issues. Lesbian and Gay Psychology feview, 2 (2): 43--9.
Espin, O.M. (1995) 'Race', racism and sexuality in the life narratives of immigrant
women. Feminism & Psychology, 5: 223--38.
Farquhar, c., with Das, R (1999) Are focus groups suitable for sensitive topics? In
R.S. Barbour and J. Kitzinger (eds), Developing Focus Group Research: Politics, Theory
and Practice. London: Sage. pp. 47--63.
Finch, J. (1984) 'It's great to have someone to talk to': The ethics and politics of
interviewing women. In C. Bell and H. Roberts (eds), Social Researching: Politics,
Problems, Practice. London: Routledge & Kegan Paul.
Fine, M. (1992) Disruptive Voices: The Possibilities of Feminist Research. Ann Arbor, MI:
University of Michigan Press.
Frith, H. (2000) Focusing on sex: Using focus groups in sex research. Sexualities, 3 (3):
275-97.
GoHman, E. (1981) Forms of Talk. Oxford: Basil Blackwell.
Green, G., Barbour, R.S., Bernard, M., and Kitzinger, J. (1993) 'Who wears the trousers?'
Sexual harassment in research settings. Women's Studies International Forum, 1 6:
627-37.
Greenbaum, T.L. (1 998) The Handbook for Focus Group Research. Second Edition.
Thousand Oaks, CA: Sage.
R E F E R E N C ES Gubrium, J.F. and Holstein, J.A. (1997) The New Language of Qualitative Method. New
an
J . (19 95) Foc us gro up S
and ethnography . Hum
Ma cDo na ld , York: Oxford University Press.
Agar, M . and
86. un, Herbert, C.M.H. (1989) Talking of Silence: The Sexual Harassment of Schoolgirls. London:
Organization, 54 (1): 78-: . darkness about life in Canada. The Vancouver S
ht m Falmer Press.
Aubry, J . (2003) City of LIg
24 February: 1 � . . Theory Jarrett, RL. (1993) Focus group interviewing with low-income minority populations:
Group Research: Politics,
er, J. (e d s ) (1999) Developing Focus A research experience. In D.L. Morgan (ed.), Successful Focus Groups: Advancing the
focus group interview� . In
zmg
and Practice. London; �
Barbour, R and Klt
ag� . . State of the Art. Newbury Park, CA: Sage.
mSh tuh' onal lma es throu h
� }I
.
Bers, T.H. (1987) Exp lon ng Research. San FrancIsCO : Johnson, A. (1996) 'It's good to talk': The focus group and the sociological imagination.
J.J. Endo (eds ), Deslgnmg an Using arket The Sociological Review, 44 (3): 517-38.
& Medicine, 17:
R.s. Lay and
Kershner, V. (1 998) Clinton faces nation. San Francisco Chronicle, 22 September:
Jossey-Bass .
es 0f d'Isease. Wo
. men talking. Social Science A1 and Al l .
Blaxter, M . (19 83) The caus .
ocwl Kissling, E.A. (1996) Bleeding out loud: Communication about menstruation. Feminism
1 ) Focus Groups in S
& Psychology,
59-69. n d Th oma s , M (200
Robson, K. a · 6: 481-504.
Bloor, M., Frankland, J., .
r: Kitzinger, C. (1995) Social constructionism: Implications for lesbian and gay
1 1 � or
Research . Lon do : Sage .
' ass et? Wom en talk abo ut the vagma.
, S . (2003) LmbTt psychology. In A. D'Augelli and C. Patterson (eds), Lesbian, Gay and Bisexual Identities
Braun, V. and Wilkmson (2» . -42 '
.

On ReVi eW,
� �
Over the Lifespan: Psychological Perspectives. New York: Oxford University Press.
.
en Sectl
Psychology of Wom mg agma e quaIs woman? On
genitals and
Wil kinso n S. (forth com .
Kitzinger, J. (1990) Audience understanding of AIDS media messages: A discussion
Braun, V . and ' . ed)
gendered identity (sub
mItt . of methods. Sociology of Health and Illness, 12: 319-55.
7.
, L . (19 98) Ou r Lab our ratings. Guardlan, 27 April.'
Brooks
1 97
1 96
 FOCUS GROUP RESEARCH

SUE WILKINSON INTER VIEWS

llenge Press, AL. (1991) Working-class women in a middle-class world: The impact of
.. e d or madness? In M . Boulton (ed .), Cha
Kitzinger, J. (1 � 9 4a) Focus gr o
d a loglca
� � � :��
p
v es in Social Research on HTV(
AIDS . Lo ndon: television on modes of reasoning about abortion. Critical Studies in Mass
Communication, 8: 421-41 .
and InnovatlOn: Metho
Taylor & Francis. s groups'. The importance of
interaction Puchta, C. and Potter, J. (1999) Asking elaborate questions: Focus groups and the
.J . (1994b) The � ethodology 0f focu management of spontaneity. Journal of Sociolinguistics, 3: 314-35.
Kitzinger, . nts. Sociology of Health and Illness, 16: 103 -21 .
between research partiClpa Sacks, H (1992) Lectures on Conversation, ed. G. Jefferson, Oxford: Basil Blackwell.
Knott, B. (1998) Where the
;
focus �ro ps �at Weekend Tele Research, Second Edition.
graph, 25 July: 7.
Schegloff, E.A (1995) Sequence organization. Unpublished MS.
Groups. A ractICai Gu ide for App lied
Kru eger, RA (1994) Focus Schegloff, E .A. (1997) Practices and actions: Boundary cases of other-initiated repair.
Newbury Park, C A: S age. . al Guide for Applied Resear
ch, Discourse Processes, 23 (3): 499-545.
Ca sey , M. A. (20 00) Focus Groups.' A Practic Seale, c., Gobo, G., Gubrium, J.F., and Silverman, D. (eds) (2003) Qualitative Research
Krueger, I,z.A. and
OakS, CA' Sag·e.'
Third Edition. Thousand ty Members in Focus Gro
ups. Practice. London: Sage.
er, R.A an d K' mg , J . A · (19 98) Involvmg Camm uni Silverman, D. (1997) Qualitative Research: Theory, Method and Practice, London: Sage.
Kr ueg . Kit Vol 4)
(The FOcuS G roup
Thousand Oaks, CA: Sage. , E ., 'Mich aels M. L., Berg,
C.N ., and Silverman, D. (2000) Doing Qualitative Research: A Practical Handbook. London: Sage.
a, M ., Sch up a k -N eu b erg ' parent'mg Silverman, D. (2001) Interpreting Qualitative Data: Methods fOJ' Analysing Talk, Text and
Lengua, L.J ., R oos elopment of a
eschl er, L.F. (19 92) U S 1
' ng focus groups t0 gm·d.e the dev ns, 41 : 1 ? 3-8. Interaction, Second Edition. London: Sage.
W Relatio
. . h , h'19h-risk fam.ilies. Family
program for dIffIcult-to_reac �� research. Stewart, D.W. and Shamdasani, P.N. (1990) Focus Groups: Theory and Practice. London:
Lunt, P. and Living � sto e, S. (19 9 6) F � S gro upS m communic atio n and me dia
Sage.
, 4, .. 78 . . . .
Journal of CommUnlCatlOn C�oyne, J . C . (19 95) Rel
ationships m Chrom'c Swain, G. (2002) Pester power. The Sunday Times, 29 December, 'Focus' section: 16.
l yons, R F., Sullivan, M.J. L., �
R1tvo , P.G., W It h Tuck, M. (1976) How Do We Choose? London: Methuen.
�

. bI'l I'ty. Thousand Oaks CA: Sage. ' bbI' tS, Weaver, J.J. and Ussher, J.M. (1998) How motherhood changes life: A discourse analytic
Illness and Dlsa . ger, G .,' Sm1' th, I,z.G ., He . nley" N. Daniels, M., Tl
. .K . and Birt D.K . (199 2) The language 0
oc h ran, S .D " Bellm f bI ac k gay study with mothers of young children. Journal of Reproductive and Infant Psychology,
Mays, V.M ., C
, O se1, 15: 51-68.
C: catlOns for AIDS risk
T., Vid oria nne, G.D. l of Sex Res earch,
our: Imph
: redudion. Journa
m en 's sexu al beha vi Wilkinson, S. (1998a) Focus groups in health research: Exploring the meanings of health
29: 425-34. ch. Newbury Par k, CA
: and illness. Journal of Health Psychology, 3 (3): 329-48.
(198 8) Foc us G roups as Qualitative Resear Wilkinson, S. (1998b) Focus group methodology: A review. International Journal of Social
. n, D .L '
M orga
Sage.
alit ative content ana1Y S1S. A g
' . uide to paths not taken. Research Methodology, 1 (3): 181-203.

rch, 3.. 112_21.

Morga n, D . L ' (19 93a ) Qu Wilkinson, S. (1999) Focus groups: A feminist method. Psychology of Women Quarterly,
Qualitative Health Resea . cing the Sta te of the Art
. 23: 221-44.
ful Focus G roups. Advan
Morgan, D .L ' ( e
Newbury Park, CA: Sage. .
d . ) (19 93b ) Suc cess
. e ,) go wrong in focus groups. QualitatIVe
. .
Wilkinson, S. (2000) Women with breast cancer talking causes: Comparing content,
biographical and discursive analyses. Feminism & Psychology, 10 (4): 431-60.
thmgs (somehm s
Morgan, D .L ' (19 95) Why Wilkinson, S. (2001) Theoretical perspectives on women and gender. In RK. Unger
Health Research , 5: 515-2 2. (ed.), Handbook of the Psychology of Women and Gender, New York: Wiley.
gy, 22: 129-52.
(199 6) F ocuS groupS. An nuaI Review of Sociolo ond Edition. Newbury Wilkinson, S. (2003a) Focus groups. In J.A. Smith (ed.), Qualitative Psychology: A Practical
Morgan, D.L. ' Resea rch , Sec
Groups as Q ua lI'ta t Ive Guide to Research Methods. London: Sage.
M organ, D . L ' (19 97) Focus
Park, CA: Sage. The Foc us Gro up Kit (6 vol
s). Newbury Park, Wilkinson, S. (2003b) Focus groups. In G.M. Breakwell (ed.), Doing Social Psychology.
Kr ueger , R . A . (19 98) Oxford: Blackwell.
Morgan, D.L. and
tive Wilkinson, S. and Kitzinger, C. (2003) Constructing identities: A feminist conversation
CA: Sag e. . sh , M .T , (19 84) Foc us gro up s: A new too l for qualita
Morgan , D.L. a nd Spam analytic approach to positioning in action. In R Harre and F. Moghaddam (eds),
. . e SOCIO . logy, 7 (3) '' 253-70. . . The Self and Others: Positioning Individuals and Groups in Personal, Political and Cultural
research. QualitatIV . h, M .T (198 5) SOCl. al mte ' and the cognitive orgamsahon
Morgan, D.L . and Spams ' .
' ractIon
ledge . S?CIOI�gY of.Health & Illn ement in focus groups.
ess, 7 (3): 401 -22 .
Contexts . New York: Praeger /Greenwood.
of health-relevant kno,,: . lOns. T OplC S and disagre Wong, G.c., Li, V.c., Burris, M.A., and Xiang, Y. (1995) Seeking women' s voices: Setting
Myers, G . (19 98) Dis playmg opm the context for women' s health interventions in two rural counties in Yunnan, China.
. , 2: : 8�-11 1 . Social Science & Medicine, 41: 1 147-57.
Language in SOCiety
g wo � en. . A ntradiction in terms. I n H . Roberts (ed.),
Oakley, A (1981
Doing Feminist Research.
) Int erv1 ewm
��
London. Routl ge & Kegan
and
Paul.
community mobilizahon
. :
S
Plaut, T., Landis, S., and Trev or, J. (1993) � ocus gr ou P
an (ed.) ' Successfu l FocuS
A case study f�om rura1
N orth Caro lma ln D . L , Morg
State of th� Art . Ne�bur Par
k, CA: Sage. ..
Groups: AdvanCing the
Rea lIty . D '
lSco urs e, !t
Rh oric and Social Construction . London
potter, J . (19 96) Representing
Sage. lligence, 28: 10- 12.
d eavesdropping . Research Inte
powney, J. (19 88) Structure
1 99
1 98
 DISCOURSE ANALYSIS AS A WAY OF ANALYSING NATURAllY OCCURRING TALK

ISS U ES IN THE DISCOURSE ANALYSIS OF NATURALLY

OCC U R R I N G TAL K

What is discourse analysis? This is a tricky question and its answers are
changing rather quickly. The complexity of this question is clearly seen by
consulting general introductions and considering the extraordinary variation
of what is treated as canonical and peripheral (see, for example, Jaworski
and Coupland, 1999; Phillips and Jorgenson, 2002; Schiffrin, 1994; van Dijk,
1997; Wetherell et al., 2001a; Wood and Kroger, 2000). One way of thinking
a about some of the species of discourse analysis is to consider them as having
evolved in the different disciplinary environments of linguistics, cognitive
psychology, socio-linguistics and poststructuralism. '
In the past, the name discourse analysis has been applied to a range of
rather different approaches to social science. In linguistics it has been applied
Jonatha n Potter
to work on the way sentences or utterances cohere into discourse. For example,
it has examined the way words such as 'however' and 'but' operate, along
with different kinds of references that occur between sentences. One of the
aims of this work was to duplicate on a wider canvas the success of linguistic
analyses of units such as sentences (Brown and Yule, 1983). In cognitive
psychology the focus has been on the way mental scripts and schemas are
This chapter urri will focus onThithes waysdis course analysishtfcanorwbeardusetasd k,toanstuddyin used to make sense of narrative. Do people work with story grammars
to understand narratives in the way they use sentence grammars to under­
naturally occof thisngchatalk . �d eembet�tbetoamstrakeaigit so. Yet I also want to stand sentences (van Dijk and Kintch, 1983)? Again, the hope was to duplicate
the course e complexp��tlesr I that m��m!eem like diversions, but if they dot innotat
WI
some of the (perceived) success of work on grammar in the psychological
explore som they are lIk� ely to remam as traps for analysts to get caugh domain.
� �1�d . Another distinctive area of discourse analysis developed in linguistics
later times. complexitie . . ?
e. Fir st , an d mo st im me dia tel y, a wI de through work on classroom interaction. Sinclair and Coulthard (1975) attempted
So what s have bs een are t1'tere he� rse analySI·s . Second, the kind of to provide a systematic model to describe typical interaction patterns in
range of thing . called. .diSCou . a broad approach to social. life tha . t teaching based around 'initiation-response--feedback' structures. For example,
discourse met I
ana1YS1S ticalw ·11 be de scn b mg IS
� assum tions theoretical ideas, analytic onentatIO�s
combines. fa-th eore course. analysldIs
and bodIes inwthe
0
is qU�lte .' Iead·.mgt an
ork . Th1· rd, itsoci mlS to think of disiol u
Teacher: What is keeping the mercury up? (Initiation)

that �l p uSh 1��:t�raldlYmaoccnyurring talogkistass wo

soc The vacuum sucking. (Response)
0
Pupil:
as a methodthat terway. Fourt t e s ?�
�
a top
our
ic
se
Teacher: Not really, Peter. Susan? (Feedback)
conceive of plex. FIfth � , I �ll cons\de� �he relationship between disc
Wl The goal here was to produce a model that would make sense of discourse
is quite comA) and convers .
atIOn a�a��:sdiSCU (CA)� structure in a whole range of different settings (Coulthard and Montgomery,
analysis (D st part of the chap these five issues as a way �f
In the .fir ome f the centraterl features of aSparti·cular discours.e ana1lyt·ic11 WI
1981).
In poststructuralism and literary theory a very different tradition devel­
introducmg (so s ? es called a discursive psychological perspectIve). oped, sometimes called continental discourse analysis to differentiate it
perspective inmet the
im
se con d half of th � � h ap ter to dis cus s an extended examprsele from its rather more straight-laced Anglo-Saxon counterparts. This is most
then move on d to illustrate the ana1y 1C mentality involved in doing discou
_

associated with Michel Foucault, and is less concerned with discourse in terms
that is intende of specific interaction as with how a discourse, or a set of 'statements', comes
analysis. to constitute objects and subjects. For example, medical discourse may
come to constitute particular objects as distinct and factual ('vapours', 'HIV+')
and the doctor as a particular individual with knowledge and authority. (For

201
 DISCOURSE ANALYSIS AS A WAY OF ANALYSING NATURAllY OCCURRING TALK
JONATHAN POT TER ., TAL K

rse see McHoul and to

DA has an analytic commitment studying discourse as texts and talk in
, '
dISCU SsI' On 0 f Fou cau lt's no tio n of dis cou That is, the focus is not on language as an abstract entity such
an accessIble social practices,
Grace, 1993), then, at least these four somewhat independent formsevenof as a lexicon and set of grammatical rules (in linguistics), a system of differences
There are, ' 1'th different d.ISC ' ry homes , To make things (in structuralism), or a set of rules for transforming statements (in Foucauldian
discourse ancat alYSIS "': rse ana1YSIS'' IS'1' PS1ma om eti me s use d as an inclusive ice 1abe1 genealogies), Instead, it is the mediumJQr interaction; analysis of discourse
ed , dIS cou ,
more complia11 f these approaches combmed wI'th speech act work, Gr IOn , an becomes, then, analysis of what people dO', One theme that is particularly
0
for some or mgUls' tic presupposIt, ion ver sat emphasized here is the rhetorical or argumentative organization of talk and
pragmati, cs, r bs, 1983; van D"�Jk' 198,5) ' ,Thcne'frou 1ca1 linguisticsincandluscon
ion here IS' an texts; claims and versions are constructed to undermine alternatives (Billig,
gh log ic of
analysis, (St 1anguage functIOn andl or a co ncern with language outside of
ub 1987, 1991),
This conception of the focus of DA may make it seem to be pitched at a level
emphasIs ond categone, s of. grammar' Phonetics and phonemics,
the restricte f thIS' Cha·pter will concentrate on yet another variant of discouentrse1y of analysis somewhere between studies of individual psychology and studies
of structural sociology, On this reading it would be an approach falling within
The rest 0, developed m' itiallY in the field of sociology and more rec ,
analysis, ThIS 0Iogy and commumc, atI' Ons (B1' ll1'g" 199 2 ' Edwards and PotterIS' the traditional remit of social psychology or micro-sociology. However, these
in social psychand M lkay, 984,' Potter and Wetherell, 1987) , The tradItion "
kinds of distinctions have been made problematic by DA. On the one hand,
1992; Gilbert led dIs,: cursi?ve psych 1ogY to distinguish it from the other
0
ion
DA has eaten away at traditional psychological notions by reformulating
noW often cal , in va rio us wa ys, D '
ISC O urs e ana lysts in this trahdit1 ogy them in discursive terms, For example, a classic psychological notion such
It IS d ' s fI C tiv e
varieties, gm�tlV, �lsm f the work in linguistics and cognitive psy� rse 0 as a cognitive script can be reworked by considering the sorts of business
0 that people do by 'script formulating' descriptions of their own or others'
reject the, co keS l't very dlfhcult proper1y to ad.dress the way dIscou1y lS' behaviour (Edwards, 1994), The suggestion is that there is a whole field of
because It rna eractional �n� lm­ �
"

act ,
IOn (Ed wa rds , )199 7), They treat the int discursive psychology which is amenable to systematic study and has hardly
to
is oriented d Coulthard (l9 5 as °ver1y based on rather mechamstIctake
of Sinclair an1 SIS' and inattenti!ve to the complex social practices thatoubts been touched in mainstream psychology (Edwards, 1997; Edwards and Potter,
guistic' an1 a Yrooms and simI' Iar 1ocat'IOns ' They have expressed similar d d by, 1992; Potter and Edwards, 2001),
place m c ass ldian approache� t? d'lSCOurse, although being impresse On the other hand, the micro-macro distinction has also been made
about Foucaud by, some of theIr mSIgh , ts, problematic, It has been blurred by three kinds of work. First, there are now
and influence a range of conversation analytic studies which are concerned with the way
in which the institutionally specific properties of a setting such as a news
interview, a doctor-patient consultation or an award ceremony are constituted
Discourse analysis
, of thl ' latt er kin d (he nce for th DA ) is cha rac ten ,
ze� by a in talk rather than being structurally determined in any simple way (Boden
se an aly sIs
Discour tlc' a1 emphas1s: on anti-re ' a1 IS ' m and constr. uctionism
, Th IS, DA
at and Zimmerman, 1991; Drew and Heritage, 1992), For example, pedagogic
meta-the, the way verSl' ons f the world, of society ' events and mn
ore 0
,
er interaction certainly happens in school classrooms, and yet much of what
zes
emphasI ' 1 wor1ds, are produced m' d'IScourse, On the one han d, thI' S lea d s happens in classrooms is not pedagogic (playing around, chatting) while much
ho iog ica om p l '
IS h ed recognizably pedagogic interaction ('test' questions, encouraging discovery)
Ptosyca con cern WIt, h parfIClp ' ants' construct IOn ' s and l'ow ' they are accconstructed happens over family breakfast tables or with a partner in front of the television,
othe�, 1't 1,eads toversioogna rec ition of the Indeed, it
and underminentd;naanturd,eonof the res ear ers o ns of the world, torical Second, there is work on the way people produce descriptions or stories
and con tin ge ve lop e� bY par;:�pants or researchers, as aetrheal., 1995; of social organization in their talk. For example, Wetherell and Potter (1992)
rea lism , wh eth er de studied the way particular constructions of social groups, processes of con­
treats
production tha t can itself be), decompose� and studied (Edwards flict and influence, histories, and so on were drawn on as practical resources
Gergen, 1994;plem potter, 199� '
IS an em ph asi s on reflexivity: fortheexaprampcticle,e for blaming minority groups for their own disadvantaged social position,
A s a com ent to thIS, th ere ' s 0fDA researchNo for That is, social structure becomes part of interaction as it is worked up, invoked
,
Im ' tlo
pI 1ca ' ns from the conclus IOn te the way I and reworked (Potter, 1996a), Third, there is recent work in discursive
what are thefor its lite rar y for ms , '
mc 1 u d 1' ng thi s ver y tex t? -
of DA an d ' h ap t er usi ng the con ven t '
IOn al homogenizing cat egoriza
, psychology that attempts to highlight the way psychological notions are
have int rod uce d th IS c , d the familiar rhetoric of progre ss, Re flex l Vlt y constructed in and for institutions, and how they can constitute some of the
C1, �htIes
'

tions of researchtospe er ::waY a text such as this is a versions,, pre selectiv:ly characteristic features of organizations (Edwards and Potter, 2001),
enc oura ges
up
us
coh
co
ere
nsId
� ce n � , oh:rence telling hist oric
mc al storie sen ,
tmg
198 9;
Typical DA studies focus on transcripts of talk from everyday or institutional
settings, on transcripts of open-ended interviews, or on documents of some
wo rki ng stltut �
mg an 0b J ec t '
IV :, ou t-th ere reality (As hm ore
and , ind eed, con kind, Sometimes these different materials are combined together in the
Atkinson, 1990; p otter, 1996a),
203
202
 DISCOURSE ANALYSIS AS A WAY OF ANALYSI N G NATURALLY OCCURRING TALK

JONATHAN POTTER % TAL K

' but prOVIde
�riented to; that is, they are not temP lates for actIOn . a way of
ly qu ali h the principled
tative, althostugthe .
sam e stu dy . DA is ov erw he lm ing
tification per ing se, but againth talk wa y counting . deviations' . TIle absence 0f a return greefl.n� do�s not dIsconfirm
mterpreting
a regularIty, rather it is the basis for inference.. the reCIpIent IS rude, sad, deaf
argument is oftnot against quan the act ivi tie s be do ne wi an d texts (see perhaps (Heritage, 1988).
and coding s en ob scu res lum e; Po tte r an d We therell, 1987;
Heritage, thi volum e; Pe rak yli:i , thi s vo So what kinds of questions are coherent w' h�' DA. GIven the general focus
?

Schegloff, 1993). IS. on . texts and talk as social practices' th::e as been a dual focus on the
practIces themselves' and on the resources that are drawnon. m. thosepractices.
Ta1�e gender inequalities, for example. Studles ' have conSIdered the way in
Discourse <lin """ '''' '''''

d is un de rst oo d as something thach,t WhICh such inequalities are constructed, , �ade factu�l and justified in talk, and
In much traditdion al soc ial res ear ch, me tho aranteeing goocasderesthaeart the they have also considered the resources ( mterpretahve repertoires', identities'
can be codifie ywiconthditspecific guidecon lines that, if noeedt gu category systems, metaphors) that are used to rr anufacture coherent and
are a necessarclusions are ion for its bydurefct.ere Ind , it is oftcorenrecthet an d complete
nce to the les, getting high levels persuasive justifications that wo k tO �.usta;. n thos: �mequalities (Clarke, et al.,
research conprocedures sucjushtifi ed ing variab in press; Gill, 1993; Ma;shall an� W et erel , 1 989, Wetherell et al., 1987).
era tio na liz
following of reliability for codasingops, and so on. DA is not like this. The any aly tic
of inter-raterused to arrive at claims is often quite different from the wa se tho talk as
procedure stified. I am going to focus here on DA specificalll .as ap ' pl�. ed to naturally occurring
claims are ju research pra ctices in the natural an d social sciences, doingingDAa , as it might
talk. However, it is important to note th t �IS.tOPIC IS n�)t as SImple
.

Like most sometimes moreJoslikh e(altsexhough,

nt of craft skiall;miitldis Ch appear. Naturally occurring talk can b: :e. ;hvely straIghtforwardly defined
has an imponrtafolntloweleingmethe ipe for skill too icken Rogan ation analysts as spoken language produced ent'Ire1y m ependently of the actions of the
chicken tha nk of it, that is rec rather a craft mentality!)., wh Convers es what is researcher, whether it is everyd�� ��n :e;satio� over the telephone, the records
come to thi k of developing ich captur formally
sometimes talher nicely (Psathas,an199ana This makes itt the clatoimdesscrareibenecessarily
lytic
ha rd of a company board meetin;� :�� eractIOn bet e n doctor and patient
in a surgery. It is natural i he SP i s:nse th�t,:It �I� not 'got up' by the
involved rattime to learn. But that do0).es no mean tha someone has learned
and it takes ate -if you cannot easily sayt pre cisely how y have
researcher using an interview schedul� eStlOnnaIre, an experimental
protocol or so�e such social research tecimoi�lgy. . .
hard to evalue, you do not have so much difficulty saying whether theequ Although thIS is useful in hi hlightin� 1l�W far tradItIonal researchers are
to ride a bik not. Likewise, there are a range of ways in which the addeviaacynt implicated in the production � 'd t ' �:� :�ggests a hierarchy moving
fallen off or analytic studies can be evaluated, including a focus on mena, fro� somewhat ephemeral interac�i�� :n l oratory to more real inter-
of discourse ng that participants' themselves orient to claimed phenontly, the actIon happening naturaIly out m' t1le world. . A better conceptualization
cases, checki th other discourse analytic studies, and, most importated with treats naturally occurring talk not as a straIghtforward discovered object'
coherence withat readers themselves can make when they are presen r does it but as a theoretical and analy(IC St ance on conversafI na I mteraction.
0 ' This
evaluation ngside of its analytic interpretations (Potter, 1996b). No ions or may seem rather abstruse but it has two advantages. On the one hand' it
transcript alocannot be learned is dependent on not merely intuit
it differentiates DA from other wor1 on records of interac.t'IOn such as content
mean that it but learning the requisite sets of skills.
-

and ana1YSIS,. which involves the kinds of coding and countmg

<:

that obscure the

imagination,ional stories of method in social research you have a questyoionu ma y subtly contexted nature of conve� f�on:l �' nt:ra�tion as well as the sorts of
In tradit for a method to answer that question . For example, rs, turn-�y-turn displays of understa �i n r baIr that have been effectively
sea rch
then you d in the 'factors' that lead to condom use in sexual en -ended cou nte use� m conversation analysis (Psathas� 1995) . n the other hand, It. provides
be intereste whether to use an experiment with vignettes, some opennking of a dIfferent perspective on research proced�res such as interviews . Instead
and ponderor DA to check them out. DA is not like that. Indeed, thi ply not of treating these as' machinery for harvestmg
interviews y can be positively confusing. Some questions are sim s about c ' as an arena for interacfIOn I. � ItS
an be VIewed . data from re 'pondents they
own right: sthat is, natural-
it in this wam a DA perspective. For example, the kinds of assumptional social
coherent fro outcomes that underpin a lot of thinking in tradition rmative interaction-in-interview. What I am suggestmg' then , is that 't IS' the analytic
1
a nd theoretical stance that constitutes its ob"J ect as nat�rally occurring talk,
factors and and survey research do not mesh with its rhetorical and nodiscrete and we should be wary of accepting too readIly. assumptlons about what kinds
psychology r than conceiving of a world of discrete variables with opted, of talk are natural and wilat are not. S ome ofthe
logic. RatheDA there are constructions and versions that may be adrmined here are usefully discussed in an excha e m
�Ul' te comp
. DIscourse ' at stake
I�x Issues
S tudzes (Lynch, 2002;
effects, in or undermined. Thus a categorization, say, may be underms are ten Have, 2002; Potter, 2002; Speer, 200�)�
responded tolarization, no upshot is guaranteed (Billig, 199 1). No
by a particu 205

204
JONATHAN POTTER TALK
DISC OURS E ANAL YSIS AS A
M

WAY OF ANAL YSING NATUR

A llY OCC U R RING TAL
.
This point is partIcularly . �rtant for showing what is distinctive about
Imp
K
1 Construction. DA is constructio
the considerable body of DA wor that h s used open-ended interviews. When on the way versions and descriptnist in the sense that it takes a specific focus
interviews are treate ? as mac hmer ' �or harvesting psychologically and construction and use of descriptions are assembled to perform actions. The
linguistically interestmg respons �s, �he research is inevitably focused on methodologists have been critic ions is a topic of study. Although ethno­
those elements 0f m ' terviews contnbuted by the participant rather than those social science, it is no t cle ar that al of the social constructionist tradition in
from the researcher. However, It. IS. POSSI'ble to conceptualize mterviews " as on construction and that in ethnothe re is so mu ch tension between this take
arenas' for interac t·'Ion between .two or more parties . That is, we can treat them some arguments, see Button andmethodology or conversation analysis (for
as a form of naturaI conversatlonal mterac ' . t10 ' analysing them the same
' n, by 2 Rh eto ric . Whereas conversatio Sharrock, 1993; Potter, 199 6a) .
way that. we might a telep�on� conversaflOn between friends or a cross-examI-. organization, DA is also focusn analytic wo rk is focused on sequential
nation m a court rOOln. Widdicombe and W ooffitt (1995) provide . one of .the versions are put together to coued on rhetorical organization -. the way
most thoroughgomg '
· attempts to use mterviews " m this way, treatmg matena 1s of sequential organization is a nter alternatives. Often an understanding
originally. collected for a study 0f socia ' I 1' dentity as examples of unf0ld'mg organization. prerequiSite for understanding
rhetorical
conversatlOnaI m ' teraction where the sense 0f social categories is refined and 3 Cognition. DP is anti-cognitivist.
reworked. Furthermore, once this perspectIve ' on interviews
. , is adopted the reference to underlying cognitIt rejects the aim of-explaining action by
standard methodology textbook mJun " �flO ns to be as neutral and uninvoIved ethnomethodology and CA is a ive states. The status of cognition in
as possible become highly probl ematIC. It only makes sense as part of the in te Mo lde r and Potter, in pr source of some disagreement (see papers
fiction that the researchers can son:eho disappear from the interacfIon 1'f cognitivist reading of Sacks. Dress). Edwards ( 995) has offered an anti­
only they can make themselves passIve e :u h in DA it has been productive practical work done by cognitiviwork is often focused on consid ering the
to be actively engaged and even ar��menta./? e during interviews (Wetherell
_

st notions (see, for example, Po

and Potter, 1992). Some of the pro em� ��:� DA work, particularly that on
Hepburn, 2003). tter and
4 Interviews and texts. Although
interviews, ar: discussed .in n l��. �t : � 03)
�� eschewed interviews, studies inCA and ethnomethodological work have
Having resIste� a �O? sImp i i c i n bet';een natural and artificial talk (Billig, 1992; Wetherell and Potte DA have often used interview material
I do not want to dImInish the �ifficulties of working with interview talk. It IS. analysts have increasingly worker, 1992). However, more recently discourse
and there are p��icular difficulties in extrapolating
contrived; it is subject to pOwerfuI expe t tions about social science research also worked more with texts tha d with naturalistic data. DA research has
fielded by participants; have been an important topic forn conversation analysts although, texts
from intervIew talk t0 activities. in ot. her settings . Discourse analysts have
.
5 Resources. DA has focused on ethnomethodologists.
_

increasingly turned away from mte �vIews to focus on materials less affected the resources drawn on in practi
by the formu 1aflOns and assump tlOns 0f the r�searcher. This. move will as the practices themselves. Th ces as well
'interpretative repertoires' (Potte is is especially so with earlier work on
undoubtedly continue as the power 0f work on natural conversatIOn becomes as to how far one thing can be r and Wetherell, 1987). Live issues remain
more apparent. continuing debate over the status studied independently of the oth
er, as the
attests (Hester and Eglin, 1997) . of membership categorization analysis
How is discourse analysi � �elated to ethnomethodo/ogy 6 Epistemology. An im
porta influence on DA has be
and conversation analysIs . scientific knowledge. This nt en the so ciology of
I have already noted th� sl?rea.d 0f different approaches that have been called truth and knowledge thanhaiss led to a more sceptical position on issues of
DA. T�ere �s some vana.tI� : �� : themselves most clearly at the levcommon in CA These differences show
to distmgUIsh the ,claSSIC t ;:thOdOlogf!al approach represented by
e other a roaches. It is common now
procedures than in the actual proel of theory and the justification of research
\
Garfinkel's Studies in Ethn?�e h°d I y (1967) and the more recent programme cedures themselves.
�
of studies of work (Garfm e , 2�0 ' In the case of Sacks's work there are
major differences between the ear1y. fo�.us on membership categories and the
In the opening parts the ch
ground issues for discouofrse apter I have addressed a numb
an aly er of back­
later work on organI�a IOn (Sacks 1992 ' see Silverman (1998) Concerns to do with transcriptio tic research. There are a ran ge of other
. conversational
for an overVIew ) . N ev ertheless there IS conSI'derable co�vergence, particularly tion, writing up discourse res n, interview conduct, COding, forms of valida­
compared WI. th, saY, claSSIC. '.SOCIO . 1oglca
. 1 work on institutions or c1 assic more detailed coverage of theseearch that there is no space to dis cu ss. For a
experimental social psy�hol�glca1 stud' � f roup processes. 2003b), Potter and Wetherel see Hepburn and Potter (2003), Potter (2003a,
. To
In mln. simplify
d, though, that these observatlOns
�
the questIon SIX areas o/. �1:er ence will be highlighted. Hold
are speciTIC to the DA tradition
Wooffitt (1995), Willig (2003l (9 87), Wetherell et al. (2001 b), Widdicombe and
this chapter I will focus on ) and Wo od and Kroger (2000). For the rest of
discussed here. the analytic mentality inv a particular example, with the aim of illustrating
olved in discourse analytic res
earch on talk In
206

207
 I

DISCOURSE ANALYSIS AS A WAY OF ANALYSING NATURAllY OCCURRING TALK

JON ATHAN POTT E R

TALK

.
h1gh . ht som e of the recurring themes in suc
l1g h work in the system (.) that I live i n .
I hop e o
addition, plonng some f the Slm
� . ilarities and differences betw CA
een (1 .0) ((smiles a n d purses lips))
0
C

as well as ex Bashir: How would a book change that.

and DA. Princess: I [1'] dunno. ((raises eyebrows, looks away)) <­
Maybe people have a better understanding (.)
maybe there's a lot of women out there
LK
OC CU R R I N G TA who suffer (.) on the same level
AN AL YS IS O F

. re�t w aYs of anal�sing discourse. It isrceuse

d1ffe ful to but in a d ifferent environment (.)

s a wi
There i . ctIOnde range of at foc us o the s dra wn
kinds of resou used . Th
who are unable to: (.) stand up for themselves (.)

d lstm ' bet we en stu dles th

m ake a ourse and the practic. es ' h' c1 those resources are hlighet
·
because (.) their self-esteem is (.) cut into two.
I dunno ((shakes head)) �-

on in dischere will be on the latter �r:m� O� s� dy' What I will do is hig ways Bashir: .hh What effect do you think the book had on (.)
emphasis concerns that analysiS. wO�kS ;ith' and one of the best the your husband and the Royal Family?
some of the. IS. to v:�rk WI'th some spec1f'Ic m aterl'als · It will try to avoidother Princess: I think they were (.) shocked,

of doing thiS l m' wntmg about meth0d that is to provide justifications to and horrified,

common goratha
er than aSSlS. t m' the conduct of analysis itself.
c and very disappointed.

aca dem ics Bashir:

(0.8)
Can you understand why?
Princess: (Well) I think Mr Dimbleby's book (0.2)
IJrJln c:e::;·:;, Diana
. a piece of talk that is interesting, and pro
wIth
'I
bably was a shock to a lot of people,

cho sen to
I have least m. Its start
. br?ad. outlIne' .to many readers . It com interviewBB
. es fro m a C. and disappointment as wel l .
(from Panorama, BBC1 , 20 N ovember 1 995 -
fa mili ar, at mte . wer IS M artin Bashir and the
rv1e ee IS see Appendix for transcription conventions)
televisi on intervie w; the
the late Princess Diana. The first thing to note here is that even a short sequence of interaction of
eteen nin ety two as
this kind is enormously rich, and could be the starting point for a wide range
1 Ba shi r: The Qu een des crib ed nin
I IS, . hh and it was In tha
t year that
-
of different studies. For example, conversation analysts have considered the
I" way the different interactional roles of interviewer and interviewee are inter­
her ( ) ann us h om ' bT
; �
An d ew Morton s book
abo ut you was pub IS h ed .
actionally produced, and the way issues such as neutrality and evasiveness
Prin ce ss: Um hm
. (nods)
you lever (.) meet Andrew
Mo rto n or are managed (Clayman and Heritage, 2002). I am going to pick up a theme
Ba shi r: hh Did
�
son allY (.) hel p him wit h
the boo k? more characteristic of DA, and discursive psychology in particular. I am going
Prin cess : I n never-
I never met him , no . to focus principally on just the two lines that have been arrowed - the two
(1 .0) . '1 dunno' s' . Why these? There are three reasons, all of which illustrate different
Did you ever (.) j2ilsl ona
'
lly assist h '1m WI th facets of doing work of this kind.
Ba shi r:
the wri ting of his boo k. First, these fragments of talk relate to broader and established analytic
(0.8)
) concerns with fact construction and the role of descriptions in interaction.
Pri nce ss: A lot of peo ple .hh h (cle
ars throat
l ife was In . (.)
The point, then, is that although I have not come to this material with a pre­
saw the distres s tha t my
� .
a sup po lve thin g
.
set hypothesis of the kind that a social psychologist might have when
And they felt (.) felt
o hel (0.2 ) i n the wa y tha
it wa s
t the y did . designing an experiment, my way into it is related to a wide range of prior
b � you r tfr ien
.
ds, >yo ur clo se fne nds
,< interests, knowledge and concerns. However, there is nothing particularly
Bas hir: id y u (.) allo
to spe ak to Andrew ° Mo
w
rton O? special about the topic of fact construction; a range of different established
Pri nce ss: Ye s I
did . Y res , I did interests could be bought to bear on this same material.
[OWhyO? Second, these fragments are easily treated as the trivial details of interaction.
If we were to make a precis of the interaction we would probably not draw
Bas hir: (.)
Pri nc ess : I wa s
(.) at the end of my tet her
I was (.) Qsl e
li.Q rat e (.) attention to them. On the video they sound almost throwaway. However,
>1 thin k I was so fed up
.
.
wit h bel ng< (.)
a s ba: sket cas e (.)
one of the features of talk that has been strongly emphasized by Sacks (1992)
seen as som eon e wh o wa
(-,l and other conversation analysts is that what seem to be its details are funda­
bec aus e I am a ver
and I kno w (.) tha t
y
cau
str ong
ses
pers�n
com plic atIO ns, (.) mental. Social scientists often treat talk as a conduit for information between
209
208
JONATHAN POTTER TALK
DISC OURS E ANALYSIS AS
A WAY OF ANA LYSI NG
NATURALLY OCC URR ING
TALK
speakers: there is a message and it is passed from one person to another, When Here is an explicit and famI 1lar
we use this picture it is easy to imagine that what is important is some basic , tere
m st to undercut a (reported) cIaIexa mple where the speaker invok
'm, The extract IS
' '

package of information, and then there is a lot of rather unimportant noise es an

programme in which the author ' from a current affairs
added to the signal: hesitations, pauses, overlaps, choice of specific words, Sa
DaVI'd Frost. Frost is asking about l man Rus hd ie i, s , b emg' m , terv
th e fa t wa iewed by
and so on, For conversation analysts this view is fundamentally misguided, Rushdie, -- the relIgI ous death sentence on
Rather than treating these features of talk as simply a blurred edge on the
pure message these features are treated as determining precisely what action
���e��:yC�h�� �:�� �ancel it now? Can they cancel it
2 Frost:
is being performed as well as providing a rich analytic resource for under­
standing what that activity is, t
-7 Ru shd ie: Yea h , b ut you kno w, the
It is for this reason that talk is carefully transcribed as it is delivered rather y wo uld , wo uld n't the
, � Ith
, out'I, going
as som ebo dy onc e sai
d ' Th e t h Ing
'
than being rendered into the conventional 'playscript' that is common in some into the kin d of arcana of IS,

:� ��
theo log y, there IS no tec
kinds of qualitative work. Note that it is sometimes complained that such pro ble m Th e ' not tec hni ca
hni cal
���
bl
transcription is unnecessary, unhelpful or even - sin of sins - positivistic! �
is that th y d O w ,
l. The pro ble m
However, it is important to remember that the potential playscript that often (Pub lic Bro adc ast ing Se
rvic e:26 No vem ber 1
passes for transcript is itself highly conventidnalized and buys into a whole 993)
Rushd ie's response to the im
f
set of more or less explicit assumptions about interaction,
The third reason for focusing on 'I dunno' is that it provides a neat way of dIs, count the claim as obviouslcla y m
that the fatwa cannot be cancel
led is to
IV
would n't they' treats the Iram, an � �ted , The familiar phrase 'they would,
·
contrasting DA with a cognitive psychological approach to talk. What might
a cognitive psychologist make of 'I dunno' s'? There are all sorts of possibilities, t1le s�rt of thing that p eople withs c1 aIm as sometl1mg ' to �e expected: it is
but one approach that might be taken is to treat such utterances as 'uncertainty of attitudes would say,' and 1' t formuthat backg round, those mt erests that set
led�e, This extract illustrates the poiates that pr�dIcta ' bT I Ity as shared' know-
tokens': that is, words or expressions that people use to report states of
to dIscount claims, ten ti a I for Inv ok ,mg
stake and interest
uncertainty, This would be in line with the general cognitive psychological
approach of relating language use to an individual's cognitive processes and � oth dis cou rse an d con rsation analy sts hav
representations (Edwards, 1997), Considering 'I dunno's' therefore has the dIf, fIculty or issue is widesvepre a d , ely e stressed that where some
th ere are lIk
virtue of allowing us to compare and contrast a cognitive and discursive procedures for dealing with 1' t F r
� exa mp le, , entothebe some well-developed
gIV
approach to talk. that exist for managing turI1 t'al(m established procedures
cedures to exist for terminal'Ing g we wo uld expec t tl1ere to be some p
, ns
One of the notable features of discourse analytic work is that the best way
(L�vm ' son, 1983; Schegloff and Sacks conversatIO and thIS" IS W hat is founro-
lytIc eXample, given that scientis ' 1973) �0{� to' take a more di�course ana­
into some materials like this may be to consider other materials or other sorts d
of findings, At its most basic, a good feel for some of the standard features ts
repertOIr, es of terms they use for ' tend t ep separate the mconsistent
of everyday and institutional talk is particularly useful for producing high­ JUst'f1 ym ' g thelr' OWn claims and underm
quality analyses (Hutchby and Wooffitt (1998) provide a basic introduction those of opponents we wou ld exp ining
for dealing with si{uations Wh e ct that some devices would be
and overview), In this case, I suggest that one of the ways into Princess Diana's ere th ose repertoires ome ogeth dev elop ed
IS, w hat IS' found (Gilbert and Mu � er; and
'I dunno's' is to consider the way issues of stake and interest have been ll ay, 1984) , �ol1owmg thIS logic, we
� thi s
conceptualized in discursive psychology, expect to find procedures that
in Extract 2, pe op �e us e to res Ist the kind of discounting t mi gh
see n
He re is a candidate discursive
Stake as a participants ' concern It w as not the product of a technique for undermmI ' , ng discounting
read'mg the newspaper and thi syste m a t'IC search rather I came across it wh '
Work in the ethnomethodological and conversation analytic tradition has In, the Guardian newspap nkin about th'IS'' I, sue, It com ile
lunacy of great artists', er hea dlin!d IPSychmtnst reveals the agony and the
highlighted the centrality of accountability in interaction, More recently, . � es fro m an art icl e
discourse analysts dealing with psychological issues have emphasized the
significance that participants place on issues of stake and interest (Edwards
and Potter, 1992) , People treat each other as entities with desires, motives, 3 The stereotype of the tor
; �
tured enl' s suffering ,
institutional allegiances, and so on, as having a stake in their actions, Referencing min d in a sea of dep res
sio n s ;� for h is art and losing his
a pro ble ms and dri nk
stake is one principal way of discounting the significance of an action or largely tru e, a psy chi atr
�
ist s YS day,
tum s out to be

,
,
reworking its nature, For example, a blaming can be discounted as merely a
Wh I' l e SC ien
' tist s, phi los op hers and p � ,
odd per son ality defect I'Itlc la�� ca n all suffer fro
, for real me n tal i nst ab m the
product of spite; an offer may be discounted as an attempt to influence, writers and pai nte rs, say ility you need to look
at
s Fel ix Post.
210

21 1
JONATHAN POTTER " TALK DISC OUR SE ANA LYS IS AS A
WAY O F ANA LYS ING NATURA
LLY OCC URR ING TAL K
Dr Post was i nitially sceptical , but having looked at the l ives of nearly ficially similar, the goal is not the ten
t analytic on� of provld . mg
--7

300 famous men he believes exceptional creativity and psychiatric occurrences of particular kinds ofcon . cou
t a lk WI'tl1m nts of
' cat ego ne s'
problems are i ntertwined. In some way, mental ill health may fuel some
forms of creativity, he concludes. (Guardian, 30 June 1 994)
A search for 'I don't know's' throu h a
04 �
ship counselling sessions provided �xtrset f � enals taken from relation­
.

The feature of the article that struck me was: 'Dr Post was initially sceptical start of a long story in which the spe akeract . . e �xtract comes from the
evening in a pub with his wife Conme. . A ' JIm ' my, IS descn'b'mg a d'Ifficult
sellor present. Ol�e of the themes m
. . . '. Following the idea that all features of talk and texts are potentially there is a coun s
. thwell as Connie and Jimmy there
to do some kind of business, we can ask why this particular feature is there. .
byJImmy that Connie flirts wI'th other men e ses sion is a sen. es o. f camplam
' ts
What it seems to do is counter the potential criticism that Dr Post is A t .
perpetrating stereotypes about madness and creativity. His initial scepticism a number of suggestions that he is p atha1o' . the same t'Ime Conme has made
encourages us to treat his conclusions as factual because they are counter to harmless sociability as sexual suggest'IOn glcally Jea ' lous and pro neIo seeing
(Edwards, 1995).
his original interests.
lar OObJ;, (0.2 )
I have suggested that such features of discourse can be understood by 4 J imm y: This tQn e particu
a medical analogy. People can avoid catching a disease such as tuberculosis any way (0.2 ) ther e was uh:
(1 .2) I d idfl' t-
by being inoculated against it. Perhaps in the same way conversationalists Qo nnie had ma de ar�
men ts to t --- mee t peo ple .
and writers can limit the ease with which their talk and texts can be under­ (1 . 8)
mined by doing a stake inoculation (Potter, 1996a). Just as you have a jab And I d idn' t l!l[an1 to. (0.6)
It l!l[as n't any Qth er thin g.
to prevent the disease, perhaps you can inject a piece of discourse to present (1 .6)
undermining. An d (0. 8) we sat in the pub
Let me now stand back and highlight two features of the kind of analytic we (.) lliarted to d iscuss=
and
mentality I am working with. First, in common with conversation analysts, =>w e had a little bit of a row
discourse analysts are concerned to use evidence from the materials as far .< (2. 0)
In the pub . (0.6)
as possible rather than basing interpretations on their own prior assump­ And illg uing about the tim
e. (0.8)
tions about people, mind, society or whatever. In this case, note that the U:m (.) whe:n the se people
cam e i n . (.)
idea that there is a stereotype about madness and creativity is not my own, >It was :< (.) Joh n and Carolin
e. (1 . 0)
it is introduced in the text itself. Moreover, the analysis does not depend And then they had- (.)
on this stereotype actually existing, merely that it is invoked as an issue in this other fella Dave.
this text. Note that this does not mean that the analysts expects to be able to °Wi th them as we ll. °
free themselves of all their preconceptions, rather it is that analysis is, to an (6 l ines om itted]
important extent, an interrogation of those expectations. they all cam � in the pub any
way. (1 . 0)
We ll (.) Con nie sat bes ide
Second, note the way I have moved in this analysis between conversational And I sat (further back).
(0.6) Car olin e.
and textual material. Discourse analysts have been much more willing than So you was (.) you was spli
conversation analysts to combine such materials and have tried to avoid The y sat in- on the Qther side
t between us .
making a priori assumptions about differences between the two. Both talk and (1 . 0)
.
texts are treated as oriented to action; both orient to issues of stake and may [1 6 line s om itted]
be inoculated against discounting. And uh:: (1 . 0)
Con nie had a sho rt skirt on
I don 't kno w. (1 . 0)
'I dunno' as a stake inoCUlation
And I knf2l!lL this- (0.6 )
uh ah- may be I had met him
So far, then, I have emphasized some background considerations that might I mus ta met Da:ve before.
. (1 . 0) Ye: h. (.)
help us understand what Princess Diana's 'I dunno's' in Extract 1 are doing. : �
But I d eard he was a bit of
(0.8 )
One helpful way to continue the analysis is to collect some more examples He dld n t care: (1 .0) who
a la:d ( ).
of a similar kind. More formally, we might think of this as building a corpus he (0.2 ) chatted up or (. )
for study or even coding of a set of data. Whatever we call it, the goal
who was in Irela nd (.) y'kn ow
chat up with .
tho se were (una vailable) to
is to help the analyst see patterns and to highlight different properties of 0 .�
particular constructions. Although some of the initial procedures are super- So Con nie stood up (0.8 )

212
213
 DISCOURSE ANALYSIS AS A WAY O F ANALYSING NATURALLY OCCURRING TALK
JONATHAN POT TER " TALK

pul led her skirt righ t up h

:r sid e (O,�) Two things are particularly worth highlighting here, First, note that despite
ight at Da, ve ( ) the various dramatic events in Jimmy's long narrative (including a suicide
and she was looking stra
,

) attempt) the very first thing that Connie picks out to contest is the descrip­
>ol ike tha to< (0,6
(D E:C2:S 1 :1 0-1 1 ) tion of the skirt length . In doing this she is displaying a skilled awareness
of the relationship of descriptions to moral categories, This is a display that
, erm ' s des criP tion of Connie's skirt lenated gth we can use to help support our own understanding of the working of this
rt by con sld ' g J '
� m m y
Let us sta , 1'mportant business rel description.
(arrow 1). For J Im ' my this descrIption does somoehas the problems that nee d Second, note that here Jimmy does not seem to be in any doubt about the
y are her e for cou �s� llm ' g, and wh It is
to why the
1 h'mg about Connie's character. precise length of Connie's skirt. His sharp, highly audible, inbreath is a display
nnie as '£lirty' an objective par cho1se
e sho rt s '
(Ir t ex em ph fles som et ,
tlcu ar
fixing, Th str uc t ' on 0 f Co of disagreement with Connie's claim about her skirt length that occasions a
I g so�ething that she
blo ck in the con �
a building
rather than just an oplm'on ' He
, IS mere1y report'n grudging modification by Connie, The point, then, is that there is no evidence
her tha n en gag " m psychological judgement. However, tely
mg
he of Jimmy's cloudy memory - no 'I dunnoness' here; precision in skirt length
to wear rat , 1 my's stake in it is , , lik now seems to be the order of the day, This variability supports the inter­
description is an especlally delicate one' wh ere J'm tIOn
for Jimmy is that the deSCrIp pretation of this 'I dunno' as a stake inoculation and is hard to fit with the
ence that he is precisely the ofthis
con cer n, Th e pro bl e
eVI�
to be a pa rti cu lar sor of plain vanilla cognitive account in which the speaker merely reports his or
be tur ned ro und an d used a s , � il
could be s every deta
' 1ous guy who 0bseS" SlVOely re' mem
her lack of certainty,
pathologica11Y Jea That " 1m d escrIPtl :��f
IS, ht gen erate problems for h'1m Let us return now to Martin Bashir's interview with Princess Diana, We
partn er's ski rt lengt h,
he mana g s delicacy? , my says are now in a better position to make some systematic suggestions about
as much as for Connie ' How can :: � � gth tha t JIm
It is immed�ately after the y gh e be saying just this just here? Let
de s ' P t' o of the ski rt len the 'I dunno' s' in this passage of talk. We can start to make sense of their role
in the management of stake and interest, and in particular their operation as
'I do n't know (arrow 2) , Wh , attempts t0 h ea d off the potential counter stake inoculations,
ady
nnie's clothing: that he was alre
sid er the P OSS I'bT 1 1 t y tha t It
US con , pectm
ms ' g Co The first thing we need to be confident of is that there is an orientation to
that Jim my wa s jea lou sly , interpr eta tion
concerned about It' even bef,ore the evenmg was under way?' This des " tIOn 0f issues of stake in this material. It is not hard to find, Bashir opens the sequence
the s �b- enc�� r �: my pro vid es a CrIp by formulating the relation between Andrew Morton's book and a hard year
IS P t of her flirtatious behthe
aIl of
is con
Conn
sis
ie's
ten t
ski
wi
rt
th
len
the
gth
det
tha t , part
IS 0 react'IOn more accountable, At mg
aviour,
, sam
e
for the Queen (her well-known 'annus horribilis'), He then pursues a line of
his
questioning to the Princess about her involvement with this book. He attempts
, ty w r1(S against the idea that he is say IS,
, in tur n, ma kes ow n stro ng th'
which . .

tam to tease out for the viewing audience how responsible she is for this (negatively
tim e the expression of uncer ?
cally Jea1ous,
h e
constructed) book.
ardly reportmg hIS
ing thi s, b eca use 1's pat hol ogI ,
, '
notic tfo rw Princess Diana responds to these questions with a series of denials, evasions,
, �, This would be in line with the
do n't kno w ' as J'I mm y tra igh
Why not treat the 'I of the nar ratlVe accounts and implicit versions of the role of the book (in that order to the first
a b ou t thi s fea tur e ty markers',
uncer tai nty
log ica l acc o � nt 0 f � ch utterances as 'uncertaindon three questions), However, having accepted that she had some involvement
cog
Can
nit
we
ive
ad
psy
jud
cho
ica te bet we en t ese d 1 � erent interpretations of 'I dis
, ht go about th'IS, One approach tha ,course t
't know'? with the book, if only via her friends, she is now faced with a tricky question
about how the book can be a positive thing (how would a book change
, US wa ys we mIg 11Y
Th
an
ere
aly
are
sts
va
ha
rIO
ve fou nd p � rti: cu r1
'!a
f 'tf 1 has been to loo k for var
bl�l�I:y � to be expected where peo
IabTt
ple are
that?), This question is so tricky because of its potential for suggesting that
Princess Diana has acted as a spurned and vindictive ex-wife, getting her
tw een dif fer ent ve rsl O ns , arI act ion s-
be , dl, ferent w ays t P erform different revenge for a book that Prince Charles had an involvement with (as she makes
con, stru cti ng the ir talk m �
, and between verSIons can be an 1' p ortant clue to � unde rsta nding at the end of the extract), The category 'spurned and vindictive ex-wife' is
" ym
se, f�r a pIe we can search the maten' aIs
varIabIlIt
wh at act ion is bei ng do ne , In t � '
IS c
, s sk�rt 1e g�� ;e d � not have to loo
� k ver y hard!
perhaps not fully available in this sequence - but it is a commonplace in tabloid
newspaper reporting, and helps make sense of both the persistence of the
oth er ref ere nce s to Co � me ' fol low ing,
for
y firs t thi ng Co nm e say s a ter JIm ' my" s long narrative is the questioning and the strength of the resistance. We can certainly identify an
Th e ver orientation to stake,
5 Co nnie', My sk'irt nro �
bab ly we nt up to abo ut the
ke of breath))
re, ((gestu res))
Given this issue of stake, we can understand the placement of the two
'1 dunno's', The uncertainty displayed in the answer to 'how would a book
Jim my: ((a sha rp inta n ev er loo ked at I
Co nni e: Ma be a bit
shorter, was done for no- , It change all of that?' precisely manages the danger she will be seen as calculating
=
--

it was my fnen d com mented

I it
part'l cular blo ke when did
lot 0' leg ton ight .
and malevolent, a woman who has carefully planned her revenge, The 'I
Oh you 're sho win g o:ff a dunno's' help break the connection between her action of helping with the
(DE :C2 :S1 : 1 1 )

21 5
21 4
JONATHAN POTTER TALK
DIS COU RSE ANA L.YS IS AS
*'

A WAY O F ANA LYS ING NAT

URA LLY OCC URR ING TAL K

book and the potentially noxious identity implied. Note the coordination of features of I don't know' discussed
verbal and non-verbal here. Her first 'I dunno' is accompanied by what might
'

talking to a psychologist, Rodge, above The sequence starts with

be called a display of wondering - she looks into the distance as if never having abou t hIS" eX-WIfe.
. Ro ss
been asked this before or had to think about it before. It is a lovely exhibition 6 Ro ss: You see tha t's wh ere you
of visual psychology closely coordinated with the business at hand. I marry her if I tho ugh t on
're wro ng! Why Wou ld
This is by no means a definitive account of the role of 'I dunno' in Extract she was a lesb ian ?
any leve l that
1 . And, of course, it does not address the very many other live and rele­ I dO � 't kno w. ((sh rug s)) Ma
vant features of the extract. However, what I have tried to do is show some ma rna ge to fail . ((la ugh s))
ybe you wanted you r
of the procedures that can help build an interpretation of a piece of discourse, Ross: Wh y" wh y, why wo u ld I,
wh y, wh y, Wh y.
and the mentality that goes with such analysis. Let me list some of these I don t kno w. Ma ybe . . .
Maybe low self esteem
features. ,,-
�
ma ybe to com pensate for
ove rsh ado win g a Sib li g
May be you w- .

Themes in the rlrrl.rlKIfSKS of discourse

Monica: W- w,-
wa it. Go bac k t o that s i b
l i n g thin g.
I don t kno w. ((sh rug s)) It's
con ceiv ab(e that
This chapter has attempted to overview some of the issues that arise when you wanted to sabotage
your marriage so the
analysing discourse. Developing analytic skills is best characterized as Sib ling wo uld feel less l i ke
a fail ure in the eyes of
developing a particular mentality. DA is more inductive than hypothetico­ the par ent s.
Ross: Tc �ow! Tha t's, tha t's ridi
deductive; generally work starts with a setting or particular discursive cul ous . I don 't feel
gUilty for her failu res.
phenomenon rather than a preformulated hypothesis. The focus is on texts
bie s', 27 J une 1 996 _
and talk as social practices in their own right. Part of DA may involve coding ('Th e On e wit h the Boo
Ross IS,
a set of materials, but this is an analytic preliminary used to make the quantity Mo nic a' brother, Rod ge
� is a psy cho log ist
b �Yfnend of Floss and Mo
of materials more manageable rather than a procedure that performs the ,
Not e, eac h I don t kno w'
nic a's frie nd.
analysis itself. There is nothing sacred about such co dings and extracts are is heavily em phasized)
often freely excluded and included in the course of some research.
DA follows the conversation analytic assumption that any order of detail
in talk and text is potentially consequential for interaction, and for that reason
high-quality transcripts are used in conjunction with tape recordings. In
addition, DA research generally avoids trading on analysts' prior assumptions
about what might be called ethnographic particulars (e.g. participants' status,
the nature of the context, the goals of the participants), preferring to see these
as things that are worked up, attended to and made relevant in interaction
rather than being external determinants. (Although it is worth noting that
this issue has become a source of not always illuminating dispute - Billig, 1999;
Schegloff, 1997, 1999; Wetherell, 1998).
DA does not use talk and texts as a pathway to underlying cognitions;
indeed, DA resolutely steers clear of cognitive reduction, instead treating
purportedly cognitive phenomena as parts of social practices. Such discursive
psychology has focused on the way participants invoke stake and interest
to understand and undercut accounts, and how such undercutting may be
resisted by performing actions via accounts that are constructed as factual.
In this chapter I attempted to illustrate these themes by way of a discussion
of 'I dunno' and 'I don't know'. I have considered only a small number of
examples. However, I hope that the insights are more general (see Potter
(1996a, 1998) for further analysis). Let me end with an extract from the US
sitcom Friends. Even with my minimal, cleaned-up transcription I think we
can start to see the way the humour in the sequence depends on the sorts of
216
217
JONATHAN POTTER " TALK DISCOURSE ANALYSIS AS A WAY OF ANALYSING NATURALLY OCCURRING TALK

politics
and theology of bottom line arguments against relativism, History of the
Human Sciences, 8: 25-49.
Garfinkel, H. (1967) Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Garfinkel, H. (2002) Ethnomethodology's program: Working out Durkheim 's aphorism New
.
York: Rowman and Littlefield.
Gergen, K.J. (1994) Realities and relationships: Soundings in social construction. Cambridge,
MA: Harvard University Press.
Gilbert, C.N. and Mulkay, M. (1984) Opening Pandora's box: A sociological analysis of
scientists' discourse. Cambridge: Cambridge University Press.
Gill, R. (] 993) Justifying injustice: Broadcasters' accounts on inequality in radio. In
E. Burman and 1. Parker (eds), Discourse analytic research: Repertoires and readings of
sis mean s doing texts in action. London: Routledge.
Antaki, c., Billig, M., �
Edwar s, D . and Potter J (2003 ) Discourse analy
. . que 0 f SlX analy t'1C Sh or' tc'om1'ngs, Discourse Analysis Online, 1. ten Have, P. (2002) Ontology or methodology: Comments on Speer's 'natural' and
analysis: A cntl 'contrived' data: A sustainable distinction, Discourse Studie�, 4: 527--30.
http : / /www.shu .ac.uk /daOl /pr vious /vl /nl gy ofscientific knowledge. Chicago: Hepburn, A and Potter, J. (2003) Discourse analytic practice. In C. Seale, D. Silverma
/index.htm
� righting s ociolo
Ashmore, M. (1 989) The refleXIVe theSls. W n,
J. Cubrium and C. Cobo (eds), Qualitative Research Practice (pp. 180--96). London:
University of Chicago Press. " . textual construction of reality .
Sage.
' n .' The
.
Atk111son, P . (1990 ) The ethnographIC Imagma t10 .
Heritage, J.c. (1988) Explanations as accounts: A conversation analytic perspective.
London: Routledge. . ology. In C. Antaki (ed.), Analysing everyday explanation: A casebook of methods. London:
' k'mg.. A rhetorical approach to social psych
B 1'll'19, M. (198 7) Argumg and. th m Sage.
C a bridge: Cambridge UmverS1't y Press . Hester, S. and Eglin, P. (eds) (1997) Culture in action: Studies in membership categorization
�
B1' ll'19, . (1 991) Ideologies and beliefs. London: Sage. analysis. Washington, DC: International Institute for Ethnomethodolo
.
Talkm g of the royaI }£ m l'Iy . London' Routledge. gy and
a h ose ord1' 11a iness? Rhetoric and ideol
Billig, M. (199 2) Conversation Analysis and University Press of America.
� ogy in
Hutchby, 1. and Wooffitt, R (1998) Conversation analysis: Principles, practices and
' M . (1999 ) Who.se terms? w
conversation a�aIYS1S, DI' O urse & Society,
B 1' ll19,
� 10: 543-58. . . applications. Cambridge: Polity.
erm a , D. (eds) (1991 ) Talk and social structure: Studies In Jaworski, A and Coupland, N. (eds) (1999) The Discourse Reader. London: Routled
Boden, D. and Zlmm . I 's ' Cambridge: Polity. ge.
ethnomethodology and conversatIOn
��r:� reement over agreement and consensus
Levinson, S.c. (1983) Pragmatics. Cambridge: Cambridge University Press.
Button, C. and. Sh�rrock, W. (19
::. �
in constructlOm st soclOlogy, J nal fior t e Theor � '
y of Social Behaviour, 23 (1): . 1-25:
Lynch, M. (2002) From naturally occurring data to naturally organized ordinary
activities: Comment on Speer, Discourse Studies, 4: 531-7.
IYSls. C ambridge'. Cambridge Umvers1ty
Brown, C . and Yule" C (1983) Discourse ana Marshall, H. and Wetherell, M. (1989) Talking about career and gender identitie
s: A
Press. discourse analysis perspective. In Skevington, S. and Baker, D. (eds), The social
Clarke, V., Kitzinger, C. and P tte , ' � : J (in ress) Kids are just cruel anyway : identity of women (pp. 106-29) . London: Sage.
:obic bullying, British Journal of SocIal McHoul, AW. and Crace, A (1993) A Foucault primer: Discourse, power, and the subject.
Lesbian and gay parents talk a ou h 0mop
Psychology. on Melbourne: Melbourne University Press.
. news interview: Journalists and public figures
ClaYman, S.B. and. Hentage, J . (2002) The . '
Phillips, L.J. and Jorgenson, M.W. (2002) Discourse analysis as theory and method. London:
Press .
the air. Cambnd ge: Cambridge Umvers1ty Sage.
(1981 ) Studies in discourse analysis. London:
Coulthard, M. and Montgomery, M . (ed s) Potter, J. (1992) Constructing realism: Seven moves (plus or minus a couple), Theory
Routledge. & Psychology, 2: 167-73 .
Drew' P. and
.
Hentage,. J.c. (e ds) (1992 ) Talk at work: Interaction in institutional settings. Potter, J. (1996a) Represen ting Reality: Discourse, Rhetoric and Social Construction. London
. . :
Cambndge: Cambr�dge Univers1ty Press .
. A study of event descriptions in conversatwn,
. Sage.
) Scnp t form '
ulatl Ons. Potter, J. (1996b) Discourse analysis and constructionist approaches: Theoret
Iogy 13 (3)'' 211-4 7.
Ed war d s, D . (1994 .
Journal of Language and Socia l Psycho .
; ical
� : 1: I t l' ns disp ositions, and rheto
background. In J.E. Richardson (ed.), Handbook of qualitative research methods
rical for
Edwards, D. (199 5) �
wo t ? tang SC 1
action , psychology and the social sciences (pp. 125-40). Leicester: British Psycholo
symmetry in relatlonsh1p trou 1es a (, �::� ; � L
r h n angu age and Socia l Inter
Potter, J. (1998) Cognition as context (whose cognition7), Research on Languag
gical Society.
e and Social
28: 319-50. .
d ' D (1997) Discourse and cogm't'10n . London
and Beverly Hills, CA: Sage.
Interaction, 31: 29-44.
Potter, J. (2002) Two kinds of natural, Discourse Studies, 4: 539-42.
d , D and Potter, J. (1992) Discu!'sive p.sychology1
��::� : : �� ��
· Londo S Potter, J. (2003a) Discourse analysis and discursive psychology.
. . McHoul and In P.M. Carnic, J.E.
ook of methods
rs1ve P Y l0gy.
S C 10
Edw ards , D. and Potter, J . (2001 ) D1scu al settin gs: A caseb
Rhodes and L. Yardley (eds), Qualitative research in psychology: Expanding perspecti
ves
in instit ution
M. Rapley (eds ), How to a �a Iyse ta lk . in methodology and design (pp. 73-94). Washington, DC:
� !
American Psychological
)eath and furniture: The rhetoric,
\
(pp. 12-24). London: Cont111u m ntern ���;� Association.
Edwards, D., Ashmor e, M. an d 0 t er, J .

21 9
218
JONATHAN POTTER ,', TALK DISC OUR SE ANA LYSI S AS A WAY
OF ANA LYS ING NATURAllY OCC
URR ING TALK

Potter, J. (2003b) Discourse analysis. In M. Hardy and A. Bryman (eds), Handbook of Widdicombe, S. and Wo offitt, R (1 995)
data analysis (pp. 607-24). London: Sage. The Ianguage of. you th sub cult ures
i� acti on. London: Harvester Wh : Social identity
Potter, J. and Edwards, D. (2001) Discursive social psychology. In W.P. Robinson and . eat sheaf.
WIllrg, C. (2003) Discourse analysis. In
H. Giles (eds), The new handbook oflanguage and social psychology (pp. 103-18). London: J.A. Smith (ed" ) Quah.tatlV. e psychology
gUide to research methods (pp 159- : A practical
John Wiley & Sons.
.
,

. , R. 0 . (2000) D omg
83)
' ' I ondon: Sage.
f s t u dymg
Wood, L.A . and Kroger -'

. d Isco
Potter, J. and Hepburn, A. (2003) I'm a bit concerned - Early actions and psychological . talk and urse ana lysis .' Met
, h0ds Jor . action
In text. London: Sage.
constructions in a child protection helpline, Research on Language and Social
Tnteraction, 36: 197-240.
Potter, J. and Wetherell, M. (1987) Discourse and social psychology: Beyond attitudes
and behaviour. London: Sage.
Psathas, G. (1990) Introduction. In Psathas, G. (ed.), Interactional competence.
Washington, DC: University Press of America.
Psathas, G. (1995) Conversation analysis: The study of talk-in-interaction . London:
Sage.
Sacks, H. (1992) Lectures on conversation, Vols 1 and II, ed. G. Jefferson. Oxford:
Basil Blackwell.
Schegloff, E.A. (1993) Reflections on quantification in the study of conversation, Research
on Language and Social Interaction, 26: 99-128.
Schegloff, E.A. (1 997) Whose text? Whose context?, Discourse & Society, 8: 165--87.
Schegloff, E.A. (1999) 'Schegloff's texts' as 'Billig's data': A critical reply, Discourse and
Society, 10: 558-72.
Schegloff, E.A. and Sacks, H. (1973) Opening up closings. Semiotica, 7: 289-327.
Schiffrin, D. (1994) Approaches to Discourse. Oxford: Blackwell.
Silverman, D. (1998) Harvey Sacks: Social science and conversation analysis. Oxford :
Polity.
Sinclair, J. McH. and Coulthard, RM. (1975) Towards an analysis of discourse: The
English used by teachers and pupils. London: Oxford University Press.
Speer, S. (2002) 'Natural' and 'contrived' data: A sustainable distinction, Discourse
Studies, 4: 511-25.
Stubbs, M. (1983) Discourse analysis. Oxford: Blackwell.
te Molder, H. and Potter, J. (eds) (in press) Talk and cognition: Discourse, mind and
social interaction. Cambridge: Cambridge University Press.
van Dijk, T.A. (ed.) (1985) Handbook of discourse analysis . Vols 1-4. London: Academic
Press.
van Dijk, T.A. (1997) Discourse Studies: A multidisciplinary introduction (2 vols). London:
Sage.
van Dijk, T.A. and Kintch, W. (1983) Strategies of discourse comprehension . London:
Academic Press.
Wetherell, M. (1998) Positioning and interpretative repertoires: Conversation analysis
and post-structuralism in dialogue, Discourse & Society, 9: 387-412.
Wetherell, M. and Potter, J. (1992) Mapping the language of racism: Discourse and the
legitimation of exploitation. London: Harvester, New York: Columbia University
Press.
Wetherell, M., Stiven, H. and Potter, J. (1987) Unequal egalitarianism: A preliminary
study of discourses concerning gender and employment opportunities, British
Journal of Social Psychology, 26: 59-71 .
Wetherell, M., Taylor, S. and Yates, S. (eds) (2001a) Discourse the01Y and practice: A reader.
London: Sage.
Wetherell, M., Taylor, S. and Yates, S. (eds) (2001b) Discourse as data: A guide for analysis.
London: Sage.

220
221
 CONVERSATION ANALYSIS AND INSTITUTIONAL TALK

reality is not confined to talk. These institutional realities also exist in and
as documents, buildings, legal arrangements, and so on. The conversation
analytic study of institutional talk is concerned with how these institutional
realities are evoked, manipulated, and even transformed in interaction.
Co nv ers ati o n an aly sis There are, therefore, at least two kinds of conversation analytic research
instit uti on al tal k going on today and, though they overlap in various ways, they are distinct
in focus. The first examines the social institution of interaction as an entity in
Analy zing its own right; the second studies the management of social institutions in
interaction. The aim of this chapter is to describe some ways to go about the
John Heritag e
second task, and specifically to identify ways of cutting into the data to gain
access for analysis. To keep things simple, I will illustrate the chapter mainly
with observations about a single "institutional" conversation which is typical
of the "professional-lay" interaction that many sociologists are interested in.
But the relevance of the entry points I describe is not confined to this inter­
action. In fact, I believe that there is a reasonable chance that they are useful
I NT RO DU CT ION in gaining access to most kinds of "institutional" data, including the new
.. . Goffman (1955 1983) established that social "workplace" studies (Goodwin, 1996; Goodwin and Goodwin, 1 997; I-leath
In a long. series of wntmgs, Ervmg mo ra I and m ' Stl ' tutl ' 'Onal order that can be treated. and Luff, 2000) that have recently emerged.
em b 0 d'les a dist inct
like othe� social in�,tltuhons, �uch;s ris
interactIon . . th family education, religion, etc. ThIs
"interac�IOn .order, �e argue ���d � oth :s a co�plex set ofsonal identity, andts
inte rac tion al righ
CONVERSATION ANALYSIS A N D I NTERACTIONAL
to face and per
and oblIgatlOns whIch are li institutions. Further, the institutional ord �r of SEQUE NCES
also to large-scale macro social I. �l sig' m. f1\ance It underlies the operatl ons
interaction has a parti�ula� SOC an� 1 m�diates the bus ine ss tha t the y CA is a field that focuses heavily on issues of meaning and context in inter­
of all the other institutIOns m SIC?Cleedty,ucatIOn aI, and lega l con duc t of soc ieti es action. It does so by linking both meaning and context to the idea of sequence.
transact. Th e por1fIcaI, econom b ' means of the practices that make up the In fact, CA embodies a theory which argues that sequences of actions are
is all unavoidably transacted Y a major part of what we mean by context, that the meaning of an action is
institution �f ��cial i �
Goffman s I ea 0� � � ���utio : � nal order of interaction" was. p�rs�Itut
t ma ke up thIS mst
ed .by
IOn
heavily shaped by the sequence of previous actions from which it emerges,
and that social context is a dynamically created thing that is expressed in and
conversation analysts who study the fracticeIs tha blish ed that through the sequential organization of interaction.
as a tOpIC m ItS ow; ight ' Con vers
. . . atIOn ana YSI'S (CA) has esta .
acqUlred Underlying this approach is a fundamental theory about how participants
the�e fr�� c�s�� l���:e com:1 and interaction, mutual sense-makin?, a��
lex and intr icat e and , in ma ny cas es,
:� orient to interaction. This theory involves three interrelated claims:
ctices are speCIal and sigm 1-
ear y m . e o . l acti .
con str uct lon pOS SI 'ble . The se pra 1 In constructing their talk, participants normally address themselves
1 y (Sc hegloff" 199 2)
social rea lIty . a l't CA studies
aus e t h ey are b � sic to hum an SOCI to preceding talk and, most commonly, the immediately preceding talk
cant bec
h w peo P 1� take turns at talk in ord
inary con-
pra ctices d scn be ° (Sacks, 1987 [1973], 1992 [1964-72]; Schegloff and Sacks, 1 973; Schegloff,
of these . on an d neg
�
otIa te ove r l aps an d m terruptions,' how various kin. ds'dofe 1984). In this simple and direct sense, their talk is context-shaped.
versah . ed and d'fferent options are activated mSI
basic action sequences are ?rganklm 2 In doing some current action, participants normally project (empirically)
fa�lures in interaction for exampnele,d
�
seq uen ces ,
' how van ous ds o f and require (normatively) that some "next action" (or one of a range of
those dealt with ' how conversations are ope
of hearing and understand'mgd- areosture �
are elated to talk; h ow laug hter is possible "next actions") should be done by a subsequent participant
sed ; how gaz e an � b o Y P (Schegloff, 1992). They thus create (or maintain or renew) a context for the
and clo
zed ; how gra mm atIc a l f orm an d dis course particles are related to turn­ next person's talk.
organi . 3 By producing their next actions, participants show an understanding of a
and so on.
taking and other mteractionalCissu . 1 esd' institutional orders in interaction. The
l!0wever, therthe e are al ? � :�
so S prior action and do so at a multiplicity of levels - for example, by an
wo rld s of cor por atl � a the classroom, of medicine, law, etc: , a�e " acceptance," someone can show an understanding that the prior turn was
SOCIal But though their realIty
{ �;
ad� ��t ��� ��n :�d th ���h talk, '1984)�
15

�voked a�dtm a lk ' t b ng" itage in interaction - their complete, that it was addressed to them, that it was an action of a particular
mvo k e d m
223
JOHN H ERITAGE TALK
CON VER SAT ION ANA LYS IS
AND INST ITUT IONA L TAL K

type (e.g., an invitation), and so on. These understandings are (tacitly) 2 Institutional interaction involves s ecia .
confirmed or can become the objects of repair at any third turn in an ongoing as allowable contributions to the bPusm l cons' tramts on what WIl. l be treated
0

3 InshtutlOn
•

sequence (Schegloff, 1992). Through this process they become "mutual . ' al. talk IS, associa ess at hand.
ted with inferential fram work ' and
contexts�
o

understandings" created through a sequential "architecture of inter­ that are particular to specific institutional s procedures
subjectivity" (Heritage, 1984).
These special features create a un'lque Iff
' " (Heritage and Gre
CA starts from the view that all three of these features - the responsiveness 1991: 95-6) for each kind of institutI'OnaI m . mtgerpr
era .mt . t b . atbatch,
cho n' the fingerpnn
to context by producing a "next" action that a prior action projected, the up of specific tasks, identities, constraints 0

:mg ma?e
creation of context by the production of that next action, and the showing of procedures that the participants depl oy an on cond.uct, and :elev�n� mferentIal
understanding by these means - are the products of a common set of socially with one another. d are onented to m theIr mteractions
Implicit in this way of thinkin g is th '
.IOn ,gen e 1 d ea that, relatIve to . ordinary
shared and structured procedures. CA analyses are thus simultaneously .
analyses of action, context management, and intersubjectivity because all three conversation institution al interac
of these features are simultaneously, but not always consciously, the objects
, <
t
range of interactional practices de 1 d erally inv" olves a red UCtlOn m . the
tion and respecification of the prac�c°!s�hab the 1?arhcIBal�ts, and � specializa­
These reductions and respecifications ar !r;;nam (Dr�w and Hentage, 1992).
of the participants' actions. Finally, the procedures that inform these activities
are normative in that participants can be held morally accountable both for 0 en expenenced as constraining
departures from their use and for the inferences which their use, or departures and irksome -- especially by the I ay partIClp � : ants (A kin
from their use, may engender. lying these ideas is the further assum ho . n tha � son, 198 2). And under­
P t, aga
of conversation, the law courts, sch00l s . mt rela' tive to the ins titution
interactions, etc., are relatively recent inv: new
ntio ns
s
t�:t�vIewave
s, d octor--patient
un d ergone a great
CONVERSATION ANALYSIS AND I N STITUTIONAL deal of social change. The institutio 0f .
is experienced as, p rior to institut� aI � ; :
0

c n ersa lOn by cO ltrast exists, and

I NTERACTION �
As CA turned to the study of talk in institutions, it began with the same
individual and in the life of the sOcie
Where would someone go in the data to
;
� . m eractlOn both m the life of the
assumptions that had proved successful in studying ordinary conversation. features of institutional interaction 100k for these and other related
? Th e s 110rt
Rather than starting with a "bucket" theory of context (Heritage, 1987) in cou
.SIX brse , is "everywhere. " But we nee d t0 start somanswer to this question, of
'
which pre-existing institutional circumstances are seen as enclosing inter­ . places to pro
aSlC ewhere and I WI'11 descn'be
action, CA starts with the view that "context" is both a project and a product be the "institut'IOna I'Ity" 0f mte ' ractlon -:' .
These are:
of the participants' actions. The assumption is that it is fundamentally through 1 Turn-taking organization
interaction that context is built, invoked, and managed, and that it is 2 Overall structural organization of the
3 Sequence organization
interaction
through interaction that institutional imperatives originating from outside the
interaction are evidenced and made real and enforceable for the participants.
4 Turn design
We want to find out how that works. Empirically, this means showing 5 Lexical choice
that the participants build the context of their talk in and through their 6 Epistemological and oth
er forms of asymmetry
talk. For example, if we analyze emergency calls to the police, we want to
be able to show the ways in which the participants are managing their inter­ I will deal with each one in turn.
action as an "emergency call" on a "policeable matter." We want to see how
the participants co-construct it as an emergency call, incrementally advance Tum -taking organizat
ion
it turn by turn as an emergency call, and finally bring it off as having been
A first thing to consider
an emergency call. is whether the interaction you are lookm . g at
Now how are we going to go about this business of digging into institutional the Use of a special turn-taking ��
of some kind of turn-takin or� i?Iza � an ' zat . . All .
lOn
involves
mteractions involve the use
talk to see the ways in which participants are addressing themselves to these
specialized and particular tasks? In general, we can look at three main types of institutional interaction �se he tlO� (Sacks :t al., 1974), and many kinds
of features (Drew and Heritage, 1992): conversation . Some, however, mv . osalme urn-takmg organization as ordinary
. tran
mations in conversational turn-taki ve very specific and systemahc sfor-
1 Institutional interaction normally involves the participants in specific goal systems can be ver irn r .I ng p�oce.dur:s. These speCial turn-taking
orientations which are tied to their institution-relevant identities: doctor � � �
they have the pote�tial �� �t� �=tu ;�n� mShtutio�a.l interaction because
recalibrate the interpretation of at p tIes opportuUIhes for action, and to
and patient, teacher and pupil, etc. I mos t every aspect of the activities that
they
224
225
JOHN HERITAGE TALK
CON VERSATIO N ANA L
VSIS AND INSTITUTI ONA
L TALK
structure. Think, for example, of how the opportunities for action, what the Overall s truc tural
organiza tion
actions mean, and how they will be interpreted can be shaped by the turn­
taking rules for interaction in a "formal" classroom (McHoul, 1978). Once yo, h�ve determined
": whe (or not) some sp eCI. al turn-ta
In conversation, very little of what we say, the actions we perform, or the orgamz ' atIO n IS in operation in o rther
overall "map" of the interactio:n� dat t�e next thing to do is to bui kino-
ld a�
order in which we do things is determined in advance (Sacks et al., 1974). In a,
This will help you to look at th �t tm.s of It� typical "phases" or "sections "
the kind s of interaction we ar: s � ?nentatIOn which is normally central in
this sense, conversations are unpredictable. In some forms of interaction
cannot always successfully be d 00 <.Ing �t. While institutional interactions
- debates, ceremonies, and many kinds of meetings - the topics, contributions,
and order of speakership are organized from the outset in an explicit and
a1ways worth making an attemptesc n bed m terms of a p11ase str
. 0 d 0 so
predictable way. This kind of organization involves special turn-taking t ucture, it is
procedures. Tl11S' IS a convenient mom ent to ' .
The decisive feature of a special turn-taking organization is that departures at during the rest of this chapter. intr . aodu ce the pIece of data that we wi11
look
a scho ol employee telephones a It IS s h ort telephone con versation in whi
from it - for example, departures from the order of speakership, or the types ch
sch00I . ThI.S conversation is d:;� mot 1er whose son may be a truant from
}
of contributions individuals are expected to make - can be explicitly sanc­
tioned, This happens in meetings when speakers are ruled "out of order," in
�
observations I wiII make are n frotn a small collection, and tnany of the
the courts when persons are sanctioned for answering when they should the m ore conversations you colle lrm ed by other caseS" m th e set. In general,
not, or failing to answer appropriately, or when children in classrooms are are studying is repr esentative (sect� th: �or: su�� ?oU can be that what you
�

. see r P r yl , IS volume). A s
punished for "shouting out" answers, or talking when the teacher is talking . �he data, you wIll ight away that th'1; c �vert satl .On IS. We turn to
In the sense of being task-focus � very
. ed' bUt ? WI also be obvi ous that no special
These explicit sanctions are very important analytically. They tell us that . "in stitutional"
turn-taking organization is mvo
the turn-taking organization is being oriented to normatively in its own right. . . I ve d m the con vers
Many of these turn-taking organizations work by specifically restricting one partICIpants' anonymity all nam .
es m th'IS conversa IOn . at'Ion . To p reserve the
c

party to asking questions and another to answering them. Interactions organ­ The Appendix to this vo'lum e outlines tl1e transcnp . tIOn have been changed.
�
ized by this kind of Q-A turn-taking organization are distinct from those, conventions .
like many professional-client interactions, in which one party tends to do most Arroyo
1 Mom : He llo.
of the question asking and the other does most of the answering. Here the 2
imbalance between the two parties is normally a product of the task the parties (0. 5)
3 Sc h: Hello Mister Wilso n?
are engaged in or some other feature of the interaction, and is not the result 4 (0.8)
of a special - and sanctionable - turn-taking organization (Heritage and 5 Mom : Uh: this is Missus W
Greatbatch, 1991).
The most intensively studied institutional turn-taking organizations
have been those that obtain in the courts (Atkinson and Drew, 1979), news
6
7
8 Mom: M
r
Sch : Uh Missus Wilson I'milson .
fr� A�royo Hig
sorry . This I'S _ISS
' h Sc ho ol ca llin g?
M ' M atai l. n

interviews (Greatbatch, 1988; Clayman and Heritage, 2002a), and classrooms

(McHoul, 1978; Mehan, 1985). As these examples - courts, news interviews, End of Section 1
classrooms - suggest, special turn-taking organizations tend to be present
in large-scale "formal" environments when (1) there are a large number of 10
9 Sch: �
.. hhhhh Was
M om: -U .... . h yes he was�artin home from school � ill today ?. _-
because uh ::h I slept fact I'm sorry 1- I ,didn ' a.' lI
11
_ In
potential participants in the interaction, whose contributions must be in I at I (' V
"rationed" in some kind of formal way, and/ or (2) when the talk is designed 12
13
well mth er. . hhhh An d Uh .
know, uh feve r:
� � �; ��
�:e f el l·ng
hh o ) h �a� uh Y.
lh
for an "overhearing" audience. However, special turn-taking systems can 14 (0.2)
be found in other contexts. For example, Perakyla (1995: Chap. 2) has described 15 Mom: this morn ing.
turn-taking practices within counseling contexts that are designed to imple­ 16 Sch: U::h hu:h ,
ment special therapeutic processes. Similarly Garcia (1991) has shown that 17 ( ): . hhh =
mediation can involve special turn-taking practices as a means of limiting 18 Mom: =And uh I don' know y'kn ow
if he'll be (') .I
k'Inna J. , st b�in laying
conflict between the participants. Finally, there are other turn-taking organ­ 19 tom orrow fer surek er n o·t
izations that order speakership by age, rank, or other criteria of seniority 20 h
aro u:n d J·(hh)u n o.w,_ He's
.
(Albert, 1964; Duranti, 1994) though, perhaps because European and North _
. ,

American societies are less hierarchical than others in the world, these systems End of Section 2
have so far been less studied.
226

227
 CONVERSATION ANALYSIS AND INSTITUTIONAL TALK

JmlN HERITAGE TALK

oriented sections, which the parties co-construct and identify as somehow
Sch: ==Okay well I ' ll
ahead en:' u : h m \90 relevant to the completion of their business together.
21
W " "
u tomorrow nig ht If we
22 Mo m : do n see 1m Identifying these main sections of the call allows us to notice other features
Sc h: I wo n' cal l yo ,
111. as well:
23 ' assume he was home
tom orrow we'll J ust
24
25 ()
,

() 1 Doing the sectional analysis forces us to see that the call is focused on a
Mo m: nnRi g[ht
26
Sch : [A:n - single topic - "dealing with Martin's absence from school." Other inter­
27 does return, actions may have more than one "item" of business to transact: a patient
Sch : S end note with him when he a
M om: 1 wHI.
28
for example, may have several ailments to be dealt with, or a family may
End of Se ction 3
29
have several difficulties that require social worker support. This distinction
can be important in analyzing institutional interaction,
30 Sch: O : kay , 2 The sectional analysis allows us to see significant stages in the parties' co­
31 M om: O kay== construction of the tasks and goals of the conversa-tions, and that for the
32 Sch: ==Than k you parties, these are incremental moves toward the completion of the business
33 M om : U h hu h Bye [bye of the call, This is significant: there are institutional interactions where the
[B' bye
34 Sc h:
-- goals and tasks of the encounter can be unclear, opaque, or even suspicious
to one or both of the participants (Baldock and Prior, ] 981; Heritage and
_ _ _ _ - - - End Ca ll - - - - - - - - - - - - - -

- - - _
- - - .� -
- - � - -

Mata , w. h0 ha s been notified by teachers that Martin

lm,
Sefi, 1992). In these interactions, the "sections" are shapeless or non-existent
In this phone call Ms, Martin's ho me to check for the parties and, correspondingly, difficult or impossible to identify
ol on the day 0f th'1S call ' calls
did not atten d sch o , ks up the phone (it could have been analytically.
b ou t s, Ma rtin 's m othe r pIC d'IVI'd ed 3 Within each section, we can examine how the parties progressively develop
on his wherea nd the call proceeds, I haveha
another relativ e o r e ven Ma rtl' r: h' Im se If) a
use, although Ms , Matalinnct c1susonters
ly (or not) a joint sense of the task that is to be accomplished and look at the
11 ' 0
e ca m t f ou r s ectI ons b eca dIstl roles each party plays in this process.
this phon act WIt, h th'IS. mother, it takes four
"

.
one pIece of busin, ess to trans 4 We can look at whether the parties agree about "where the boundaries
of activity to achIeve: are" as they shift from one section to another (Robinson and Stivers, 2001).
, t sectIo ' n (1ines 1-8 ) is an "open
ing" section in w�i:h the In this call, the parties make very "clean" transitions from one section (and
fIrs for
1 Opening: The
ter int o a stat e of , teractIO
m ' n and es tablish their identItIes one component of their "business") to the "next." But confusion and foot­
dragging are also possible: one party may want to move on to the next issue
parties en
(Sc, hegloff, 1986). Matalin gets to the while the other party is reluctant to quit the current one, Or one party may
one another " e sec on d sec ' n (1'mes 9-20) " Ms
tIO
' 0f Martin's absence,, an
I n th d the not recognize that a "next issue" is now relevant, while another is trying
'
2 Problem inzt Iatw n:
11
" smg the quest IOn
"business " 0 f t h e ca by ral m initiatio n sect IOn, to press on with it . Different interests (and clear conflicts of interest) may
' th e "pr oble ' p1e
,
IS
m other exp1 ams OIves the problem in this ca11, sIm
1't , I hav e ter me d t h be involved in these clashes,
mot h er res
do not a�ways. happe�· Ms. Matalin details the
because alth oug h Marti n's
resolutions of this kind saY 2 1 9) , Using this four-section framework, it is relatively easy to identify the same
' pO ' sectIOn (l Ines Marti absence in the hght of
.
3 D isposal: In the "d IS e will ta k e tow ar d n's sections, occurring in the same order, in most of the phone calls Ms, Matalin
action sh uld take
bureaucratic
oth er , s account, and descn, bes the act'10n that the mother sho makes. However, the purpose of identifying these sections is not to exhaus­
tively classify every piece of every one of Ms. Matalin's interactions in
the m , what the child should do as well.
In other calls, �he descIon : lbes
(lin es 30- -4) is devoted to managmg
, these terms. Still less is it to assert that these sections will always occur in her
4 Closing: The fmal sect
of the cal l
t'10n (Schegloff and Sacks, 1973), conversations in this order, or even that they will always occur (d. Byrne
d' Ina ted e xit fro m th e con versa and Long, 1984 [1976J; Robinson, 2003), In other cases of these school calls,
a coor
, . ' ed th se four sections, we can see that each of we can find the participants reopening sections and reinstating task orientations
Now th at we
the
h ave
pursU
Ide
l't
ntlfl �
0 f a sp eC!f'IC sub
-g oal . Each section is jointla ytask to that they had previously treated as complete. So we are not trying to find
invariance or even statistical regularity in the presence or ordering of these
involve s involving ,
d. bY b O th P articipants as t satisfactIon of sections. The purpose of describing these sections is to identify task orien­
to -, indee d co-constructetask
d . I n th' IS ca 11 , all s are fuIf'l1 led to the ap pa ren , l to tations which the participants routinely co-construct in routine ways. Overall
not alw 0�11;�en and itntify is not essentla
achIeve
parties, but thi� do�snal ta l k , a are ide ing h ere are goal- or structural organization, in short, is not a framework -fixed once and for all - to fit data
section s of inst ltu tIO
229
228
JOHN HERITAGE TALK
CONVERSATION ANA LYSIS
AND INSTITUTIO NAL T
ALK
. . Rathen't IS' something that we are lookmg. for and I00kmg
' a t only to the extent
mto significant because, given tur
. organIzmg
that the parties orient to it m . . thelr' ta Ik. in play in this interaction), the
a sen
n-taking system for conversation
(which is
tenc
could intervene with a question or e boundary is a place where Ms. Matalin
Sequence organization explanation that the m other is piecina new observation, and thus disrupt the
talks in the way she does so as to g together. It seems clear that the mother
With the third level of analysIs. - sequence organization - we come to a :ery that she does this so that she can avoid creating thes e opp ortunities, and
central aspect of CA work. It is by :n e�ns ° f pecific actions that are orgamzed not called the school without beinconclude her explanation for why she has
in sequences that the participants 1m. tlate" d :velop and co�clude the business completed her explanation that she g interrupted. Thus it is only
after she has
they have together, and genera11y mana.ge their encounters. In analyzmg ' she takes a breath at a sentence bounhas not been feeling well "either," that
sequences, we essentially look at how partICUlar courses of action are initiated by the four h's !). dary (and it is a big breath as ind
0
and progressed an�, as part f his how particular action opportumtles . . are
If the m other's talk to this p oint
icated
:
opened up and ach:,�ted, or wi hh�ld from and occluded. All of these POSSl­. progress, her subsequent elaborati is managed so as to retain the turn in
bilities, while expllCltly analyzed by. �s, a e also implicitly grasped to a to relinquish it. Extending her turon seems to emerge because she is unable
greater or Iesser extent - by the partIclpan�s who may use what transpires
_

child's illness, she p auses at line 14,n at lines 12-13 With a descrip tion of the
as a basis for inferences about the character and situation f thelf. co- 0 In response to this, she recomplet only to find no uptake from Ms. Matalin .
es her turn with an incremental
interactants. dant) time specification ("this mor (and redun­
Ms.. Matalin's phone ca11 IS. a nc. h source for sequence analysis, but here from Ms. Matalin ("uh huh ") that ning"), and then encounters a respons e
.
we WIll Just f ocus on . one 0
aspect f the conversation: the fact that after standing that the previous speakeris prototypically used to indicate an und er­
Ms. Matalin's questIOn at l . � 9, the mother's reply seems to go on and In the face of this response, the (in this case, the mother) is not yet finished.
on and on. This an accomp;IS�ment. If we look at the structure of the reply,
. htisaway son's condition (lines 18-20), finalmother continues with a prognosis of her
we can see ng that the mother answers Ms . Matalin's question in the Thus in this exchange of ques tion ly coming to a halt at line 20.
very first line of her response WIt. h "U"' "' h yes he was": see that the mother treats Ms. and answer-pIus-elaboration,
a fault in her conduct, a fault for Mata lin' s question as implicitly poinwe can
9 Sch: �
[ . hhhhh Was Martin home from school ill t d y? = is at pains to supply an explanatiowhich she is accountable, and for which she
ting to
1 O Mom.' U " ' h yes he was * In. fact * I ' m sorry I dldn
1_
. �
ca.1I * n (for some parallels in medical
�e��use uh::h I slept in late * I (.) haven' been feeI'ln� see Silverman, 19 87: 233-64,
Her itag e and Sefi, 1992, and Heri encounters,
�� well illt her. . hhhh And u h .hhh (0.5) u : : h he had u h yl h
Lindstrom, 1998). Her treatment
but rather embodies a particularof the question is not as a "casual inquiry,"
tage an d
13 know, uh fever:
standing of its relevance. Subsequ- and specifically "institutional " under­
14 (0.2)
her account as the p roduct of ently we can see the further extension of
_

1 5 Mom: this morning.

16 Sch: U : : h hu:h, will make the convers ational an implicit sequential negotiation over who
h: h� the mother's rather lengthy turnrunning . The detailed internal structure of
� Mom: � u h I d o n ' know y'know if he' l l be (.) i � .
1 7 ( ):
negotiation. There are many is thus the product of a complex sequentia l
� ;o�orrow fer sure er no:t, He's kinna j'st bin laYing exchange that merit anal Sis ofother aspects of the sequences making up this
20 arou:nd j(hh)uhkno:w,= production. We will catcYh som the "institutional" relevances that inform their
area where initial anal Sis mi e of these aspects as w e move on to the fourth
' " (to
However, the mother th en contmues . Wl'th an apology for not "ca11mg Y ght proceed: turn design.
.
notify the schooD, and then WIth an elaborate series of explanations for the Turn deSign
situation. A noticeable feature of her turn fro.m rne 10 to line 12 is that she .
. d anes. �t all the other points where
15

careful to avoid pausmg at sentence

. 11y ob oun Turn design is an
her sentences are grammatlca 1� te (marked with an asterisk in thO a ction . When we talimp ortant place to examine the "in
stitutionality" of inter­
transcript), she (1) is ca:eful to avO!.� n;�l ("final") intonation (which would k abo
15

distinct selections that a ut a turn being "deSigned, " we are p ointing

be marked with a penod - see the ;ra�s�f�t conventions in the Appendix), talk is designed to p erfo p erson's speech embodies: (1) the action thatot two
and (2) moves straight to the next sentence Ithout a break It is also noticeable, the action (Drew and Herrm and (2) the means that are selected to p erforthe
looking at line 1 1, that she only p;uses � mmatical �laces where she is One sense in which a turnitage, 1992) . m
�
unlikely to be interrupted a�t�r e � '�:cause" in line 11, and also aft�r which someone wants is "de sign ed" concer the selection of the
to accomplish in a turn atns
she starts a new sentence WIt t e :�:d "I" (also in line 1 1). All of this IS
actio
on health visitors' home visits to mothers of new talk. In work with Sue Sefni
borns, I came across the
230

231
 -

JOHN HERITAGE TALK

following sequence in the health visitor's opening visit. The father and mother
CON VERSATIO N ANA
LYSI S AN D INSTIT UTI
NAL TALK�
4
=:
(1 .0)
respond to what looks like a casual observation by the health visitor by
performing quite different actions: � �: Yeh . They [learn so
quic k do n 't they.
[We have noticed
hav' n 't w-
7 HV: That's righ
' t.
1 HV: He's enjoying thgt [isn't he.
8 F: --> We have n oticed (0 .
2 F: -) [oYes, he certainly is=o 9 8) ma k Ing
' a grab for y
(1 .0) our bottles.
3 M : ---> =He's not hungry 'cuz (h)he's ju(h)st (h)had -, 0 F: H m[:: .
4 'iz bo:ttle .hhh
5 (0.5)
11 HV
.
[0oes h e: ( ) Ho : �; �
te does he go bet
ween his feeds?
6 HV: You' re feeding h i m on (.) Cow and Gate Premium.= ( v. A 1 .2) (Drew
and Heritage , 1 992:
Slgmfi. cantly, the m other "s resp
(HV:4A1 : 1 ) (Heritage and Sefi , 1 992: 367) , . 34)
.
The health visitor's remark "He's enjoying that" notices the baby sucking the heaI th visitor:s remark onse aVOIds the "exp�rt-novi�e" stance tha
agreement(s) (at hnes 6 and mig ht be seen as expressmg, t
or chewing on something. (Unfortunately, we do not have a video tape, but 8) seem d s" d to pr ve t WhIle the fat her' s
certainly this is how the mother understands the word "enjoy" when she
responds "He's not hungry . . . " (lines 3-4).) In replying that way, the mother
�hat t�ey are observant and alert abo
I� ��Ich they design their a u :��� ne v bab
� ? ? the health visitor
y. fhe d ifferent ways
:-
treats the health visitor's remark as implying that the baby is "enjoying" dIVISIOn of labor in the familyctions rnay also pomt to the sam e underlyi
whatever he is sucking or chewing because he is hungry - an implication putatively the junior partner in th:t : �u�ges,ted earlier. The father, whong
which she rejects by observing that the baby has just been fed. The mother's eager to p rove their competen th f . i .Y s chIld care arrangements appears is
ce m'
ior. The mother's agre em t, by notIcmg the details of their child : s behav-
response, then, is "defensive" in rejecting an unstated implication which contrast, see ms t0 aVOI' d tak
she treats the health visitor's remark as having conveyed. The father, by and '"mexpert" position ofen"pro ing the "in or"
contrast, simply agrees with the health visitor. asserts an agreement in blan ' " anything to the hea1th viSitor, butferi
v mg
rather
Thus in "constructing" their responses (quite apart from the particular The alternatives that may bdean . doglener.al terms.
contrasted as they are here mv ved . m tUrn de " are rarely as e
designs of their turns), the mother and father have elected to perform alter­
in the same responsive where d ' fferent speakersSIgn
1
xplicitly
emp1oy d'1 fferent deSign
native activities. Both activities, of course, have a "logic" as relevant next p osit ion . M o re usually, w e analyz e turn s
actions. The father treats the health visitor's remark as innocent while the looking at the details of a turn' des ign by
intera�fIOna1 p urpose or sign s co mp one nt features an d by determ
mother finds in it an implied criticism regarding the proper care of her baby . ' ining their
They thus construct their responses differently by selecting different "next" �o Illustrate this, I want to go ificance
'
at lme 9: "Was Martin ho back to Ms �atalm . ,s
telephone call and look
00
actions. These two actions may well reflect a "division of labor" in the family, me fro m 'l
frequentlY begl11S . sc h III tod ay?-. - " . M s. M atalm
a:�TI;s she calls with this
in which the mother is treated as having the primary responsibility for her ' her l11quiries to the £ ' very
baby (reflected in her defensiveness), while the father, with less responsibility, an important sense, it is a , -
que stio n. In
nearly exactly) OVer and ove "highly desig e turn that
can take a more relaxed and "innocent" view of things. is rep eate d exa ctly (or
One way of analyzing a r again. .
The second aspect of turn design is that speakers also select among alter­ the interactional hi hI
tingenci�s �t Ig
des Ig� e �
. t turn of this kind is to think of
of Ms. Matalin's callcon s, there are tw� mr::m P OSSI� . a.d,dressing. In the context
native ways of saying something or performing the same action. The following
extract - from the same health visitor interaction as the previous one - illustrates turn up: Ilihes that her question mig
this clearly. In this extract, the mother and father each perform a broadly similar ht
activity - agreeing with the health visitor's suggestion that they will be
"amazed" at the child's progress (in physical development), and they do so The child is away from scho
nearly simultaneously (lines 5 and 6). But they design their agreements rather ��eurpse, is bad news in its OWnolright sick and the r nt knows
chh� �who may be m
it. ,Sickness, of
arent (who may be worried for the
differently. While the mother's agreement refers to the development of children But M s. Matalin's call is and 1os;ng pay by being awa pain) and
.
in general ("They learn so quick don't they"), the father refers to their als o bad news or the p ar nt at y from work)
pare�t of a sick child is
experience of their own child's progress ("We have noticed hav'n't w-"). While the sItuation, There is if nor mall s�p:s;� o call 111 to notify r level. Th�
� : ano the
the father's utterance exhibits a commitment to noticing their own child's and school such that th yo u lik Ia 0 m o:mal contract betthwe school of
behavior and development, the mother's response does not.
the child is miSSing, bu e sch ool t�lIs t�e pare�t If they have een home
i HV:
2
--> It's amazing, there's no stopping h i m now, you ' l l be
amazed at all the d i[fferent things he'll start doing .
; he school if they suppose t the p are t qu� ly have the obl
�ct t�at Ms, Matalin has had ttothe ��l� ;llIf not be coming to school.toThe
tha
reas
iga
on to believe
tion tell
o a lIttle "breach of c call in t e
ontract" on the parent}srst
3 F: [(hnh hn) p lace may be the product
p art. It is just that brea ch of
232

233
II

JOHN HERITAGE TALK

CON VER SATION ANA LYS IS
AND INST ITUT ION AL TAL I(
contract which the mother's defensive explanation, which we looked at Lexical choice
earlier, seems designed to address.
t0 I�sh' �utl. Onal tas�s an d contexts
2 The child is away from school and the parent does not know it, i.e., the child A clear way in which speakers orient .
f ter m�. or � n �tance, whIle som eon e
is truant. For this possibility, Ms. Matalin's call may involve a very serious through their selection of descrip .

is
use "cop" in ordinary convers tive
piece of information for parents who, up to now, may have no idea that
their child is not attending school. �re likely to select "police officer�' ����e:d ac�s,mg
�; gIv eVI. dence in court theyt mi gh
mvolve selection is evident wh �� 1979). The fac that this can
off the beginning of "co p" ("ken e� kers - as m Jefferson's (19t74)
It is in this context that we can begin to see that how Ms. Matalin opens up
stud·Ies that have dealt with theucon h. ) m favour of the word "pol'Ice.data - cut
the topic of this call matters a lot, and that her opening utterance involves quite .. . . " M any
speakers select descriptive terms wIt1ext .�h-serenSltlvlty of descnpf IOns s110W that
a bit of caution. � fItted to the institutional setting,
Consider the turn itself: (1) it indicates that the child has not been at school
"today" (i.e., on the day of the call), but does not assert it as a fact: that the
or their role within it (Drew a d
. g as 9
��
illustration - first noted by Sack� (19 � age, 1992) . . A dramatically clear
r
child has not been at the school is presupposed in the design of the question speakm a member of an or�ranizat'lOn, 9per 64_72]) - IS the way that, when
sons � ay refer to themselves
as "we", not "I" (Drew and H r t �
rather than stated as such (Pomerantz, 1988). (2) It offers as an account for the
(lines 23-4). Here Ms Matalin mI\la � 992 ) . Th ere IS. a clear case in our data
decision ("1 Won' call' you tomorrow1ry . escnbes a course of action as her OWn
child's presupposed absence the most commonplace and the most legitimate � .
reason for the child to be away from school sickness. (3) The question is
designed so that the easiest response for the parent will be an affirmative "yes" will be made if Martin is not at sch00I mght")' bu t then adds the inf . ere nce that
'11 '
tomorrow.' "if we don' se� ,1m
response to the possibility of sickness. In CA terminology, the question � e Jus t � ssume he WaS home ill." . Here the "we " referred to as �orro:v to
"prefers" a "yes" response. (4) Even if the child is, in fact, a truant, the inquiry mference IS eVidently the school as an ms " tlOn .
' tItu
.
makmg thIS
avoids any implication of this and particularly avoids any accusation of Another systematic type of lex'Ica 1 seIectIon mv
. . olv
"in stit uti ona l eup hem ism . es what might be termed
truancy. (5) The question does not in any way directly thematize the parent's .mstItu
. tIOn . " H ere Issues that ma y be probl em
responsibility to inform the school. Instead, it leaves it open for the parent,
. 's
representatives to address for om e ahc. for the
rea
where relevant, to assume that responsibility as our mother in fact does. This, the New York Times (November 5 199 . �
�� son are do wnpla . In
It IS reporte� that Mi crosofyed
then, is a highly judicious, cautious, and "institutional" piece of question giant software corporation no
but rather of "industry leadershilo�"gIer � t? talk of md . ust t - the
euphemistic a patient will be a�ed �,;n� Icme;, references to pa in are often
design (see the next section but one). ry dominance"
Now in recognizing that this is so, you do not have to assume that Ms.
Matalin is a very tricky, Machiavellian type of person. You just have to remem­ (Heritage and Sorjonen 1994) In Oth �I sore? rather than "Is it painful?"
at, when Ms. M atalin 'does � t ter p one calls like the one we are looking
ber that she makes dozens, even hundreds, of these calls every week. She has
learned the range of possible responses that mothers make to her question, absence, she often says "We ne�d t� �l� adequate �xplanation for a child' s
up ." While this seems to indicate 'u m er to come mto the office to clear this
tio� �f :he �hIl. �s ac o��tab�I�tIty
and she has also learned that certain ways of asking this question can attract t matter f bureaucr �c record keep ing,
resistance or cause arguments. So, for Ms. Matalin, recurrently raising this
topic is like a "wind tunnel" experiment: the "wind tunnel" of repetition leads
it leaves op en the broader
and what kind of punishmques �
ent mig ht be mvo ved �m cleanng for the absence
record. up" his or her
her to a question design that evokes the least resistance. You can see this
wind tunnel effect in many other kinds of institutional talk - in medicine, Lexical selections can sha
pattern of the interaction. Forpeexawh ole s;�u�nc� s �nd, WIt. h them, the ove
social security offices, emergency calls to the police and fire departments, and call runs as follows: m p Ie, e egmnmg of Ms. Matalin's rall
others - where the institutional representative has a repetitive set of tasks to phone
be worked through. 1 Mom: Hell o
Thus, the second sense in which one can say that a turn is "designed" is 2 (0.5)
that there are always alternative ways of saying something from which 3 Sc h: Hel lo M ister Wil son
?
speakers, unavoidably, make a selection. The syntactic, lexical, and other 4 (0. 8)
(e.g., prosodic) selections by a speaker are aspects of a turn that articulate with 5 Mom : Uh: this is Mis sus
Wil son .
the performance of organizational tasks and, very often, are shaped into "least 6 Sch : Uh M issu s Wil son
I'm sorry. Thi s is Mis s M
7 at aI 'In
resistant" forms by the repetition of those tasks. An important component of 8
from Arroyo Hig h Sch ool call
ing ?. -
Mom : Mm h m
turn design is our next topic: lexical choice.
At line 8, the mother, rather
Which is the kind of action than gr��hn
. g
Ms . Ma�ahn. by
saying "hello"
that norma y occurs at thIS pom
. t (Sc hegloff, 1986),
234

235
 CONVERSATION ANALYSIS AND INSTITUTIONAL TALK

JOHN HERI TAGE TAL K

egloff, asymmetry - between the speaker and the hearer of a turn at talk; between
- a p rotot yp i cally non- committal "continuer" (Sch the initiator and the respondent in a sequence of interaction; between those
just says "mm hm " , to pr eed ith the conversation w one , No
1982) that inv ites Ms :� atal m , n��y ev:n a hostile, action from someone
�
who, more broadly, are active in shaping topics and those who are not; and
e th at thIS IS an unfr ' look back up the between those whose interventions are decisive for the outcomes of conversa­
could imagin ing to s c ° o 1 ? fflCl' als But if we tions and those whose interventions are not (Linell, 1990; Linell and Luckmann,
who does not like talk , ve baSIS for th� mother' s action, In particular,
sequence, we can ,
s ee an alt erna tI
herself using a particular, lexius ca1, ch Ol' �e 1991), From this standpoint, the contrast between the symmetry of ordinary
, Ma talin i den tifies conversation and the asymmetry of institutional discourse is indeed over­
we can see that Ms
" zatlon�1 'd" identific
ation By mg thIS
le, �anc
me plus org am simplified (Robinson, 2001; Stivers and Heritage, 2001): all social interaction
a formal "las t na Matalin" or just "Nancy,"
identifi cati on , rath er than , for �: am ? ,, !n
ess call d , specifically,ear a "call about must inevitably be asymmetric on a moment-to-moment basis and many
' w1len,
, th e p hon e c all a s a b usm interactions are likely to embody substantial asymmetry when moment-to­
she Identl , les
'f'
" (1 n fac t , tha t pro ces s ' to emerge even lier
begms r usmg, moment participation is aggregated over the course of one or more encounters ,
school bus m ess, Mata lin nam es the mot he
trying to identify who shtha e IS, talkm ' g t�' M s, e mi, staken Yet at a more general level, it is clear that there is a fundamental distinction
1 " ra ther n a mo re mformal identification - ththe VOlce ' between the symmetry of ordinary conversation and the asymmetries of
"Mister W
lO
t'
'l son
n (Mis ter for MISSU' S ) see ms to arise because the mo ndrs st0 M s, institutional interaction, The general operation of ordinary conversation is
idenh'f'lca p,hone ) S� ; hen the mowith ther res po not tied to any particular set of social roles, identities, or tasks, If it were,
sounds rather deep �n the ss- one n te ci i
se -i e ntific atio n "mm hm" at line
conversation would be a much less flexible and sophisticated institution, In
Matalin' s fo rma l, bus me , to proceed with the business-based "reason
't' Ms Ma tal m many forms of institutional discourse, by contrast, there is a direct relationship
8, she is in f a ct mVl
' mg ' , has dearl P rojected right from the start. That
talm
-
between institutional roles and tasks on the one hand and discursive rights
Ma
f 1't " t:e very brevity and theco nomy of this
for the cal l" tha t Ms,
n - an , bec ause o rliest stages and obligations on the other, For example, institutional representatives
1 eXlca1 selections made
clear , proje c tio � at e ea
eq ue . nce - ans es fro m th e commonly ask questions and require of lay participants that they answer them,
openmg s In this way, they may secure the initiative in determining (1) when a topic is
of this telephone call, satisfactorily concluded, (2) what the next topic will be, and, (3) through the
design of their questions, how that new topic will be shaped (Mishler, 1984;
In tera ctional as
ymme tries ,
mm ,
' mg Drew and Heritage, 1992), Thus institutional representatives can often direct
, ce at wh ich to be gm exa
ctIOnal asymmetrie s are a pla ntion four the interaction in ways that are not found in ordinary conversation,
Finally, intera , ality of
, tera
m ctl' Ons , Here, I, will briefly me
the s peCif" ic m '
stlt u t'
I on " at'lOn, (2) "know ho w" about the
types of asym me t ry that i nvo lve ' (1 ) partICIp
" w h'lC11 1't is embedded; (3) kn ow led ge; and Asymmetries of interactional and institutional "knOWhow"

tion and the , ltutlO

mst ' ' n m An important dimension of asymmetry between the participants in insti­
interac tutional interaction arises from the difference, and often tension, between the
(4) rights to knowledge, organizational perspective that treats the individual as a "routine case" and
? � ies of par­ the client for whom the case is personal and unique, Ms, Matalin's phone call
racf10n document asymmetrinsti
pa rtici at on
Asymme tries of ,
stu ,
dies of mstl tutI Ona l mt � that tutional to the mother in our data was one of around a dozen she made that day and,
Many ter actIOnS ' and I' n particular involving doctors, for her, it was absolutely routine , For the mother, however, it was an unusual
,
in
ticip, atio n in inst itutio nal
' 1ay-professional
encounters - for example' these mt " eractIOnS and morally threatening occasion, The parties, therefore, brought asymmetric
partiCipants m tai n the ini tia tiv e in
' 1 workers' etc ' -- tak
e and re ' nS experience and reasoning to the encounter , All agencies have procedures
teachers, SOCIa , ler 1984 , F ran ke l 199 0), Un de rlying these observatIO for the routine management of multiple cases, for "processing" cases by assign­
(Linell et al., 1988; Mlsh wIth ', a s and ar'd of "equal participation" between
'
ing them to routine categories, and so on , However, the clients -- whose
is an implici t contra st �
A s - L'me11 and Luckmann (199 1) have
, con ve rsa tio n n the inquiries, troubles, illnesses, claims, and the like constitute an organization's
speakers m or d'mary be cautio This dichotomy betweeours e
to " us a b ou t this ' routine cases - may not be really aware of, or concerned with, the pattern
commented, we needrsatIO , n and t1le asym , metri es '
of institutional dlSC h into which their individual cases fit. The client's perspective often arises
symmetries of conve ture f as,y�metry and overlook the ways in w e hic
can oversimplify the nver na ther out of the particular circumstances which bring him or her into contact with
satlO
?
n can e aS Y m � etric As they observe: "if eva the organization, perhaps for the first or only time, or at least not frequently
talking in ordinary co
'
ne s at all bet ween peop l e, l.e, if �ommunicatively rel neentd enough to have developed a self-conception as a routine case, In doctor-patient
were no asymmet er� n��;-��:�:�� ' f there would be little or no ewed encounters, this gap between routine institutional "knowhow" and singular
inequalities of knowled geuDlv:ca tlOn , nd Luckmann, 1991 : 4) , Vi at experience can be extraordinarily stressful (Zola, 1987; Whalen and
fo r m ost kin ds of co mm
rson ; P artici ate in talk and to wh o f
n can �mbody several kin
ve that ask s whic h pe ds Zimmerman, 1998) and can emerge in behavior that can be experienced as
from a perspecti or ,
dmar y convers a lO
effect, it is appar ent that
237

236
 CONVERSATION ANALYSIS AND INSTITUTIONAL TALK

JOHN HERITAGE TALK

5 Sch: .hhh Well he was reported absent from his lbjr:d and

1996 ). T .
1lIS ga can .exist and be
significant, in all forms 6 his fifth period cla:sses today.
ard, . tn c a�d social service encounters,
very callous (Mayn
k. In som e psy c� 1a
ness 0f the professional obje
of institutio nal tal ctives
. Here Ms. Matalin does not say "he was absent from . . . "; instead she says "he
may h ave o nly a dim aw are ge
the "client" . . ldo ck an d Pri or, 198 1; nta
He wz reported absent from . . . . " By including "reported," Ms. Matalin invokes
acro ss th e enhr e enc oun ter (Ba t 1
. dea
. 1995). I h th 1 caller may ha
being purs ue d ve an exa c an (unnamed) source for the information and thus portrays herself as relaying
and Sefi, 1992; Perakyla, nver
se of the co sa �
t� o �: �' m:y � e un ab le to gra sp the p.oint of
Istan ce
the information she is giving. She thus avoids underwriting the information
of the purpo . n. For e xample In . a not.on' ous call for emergency ass
' . . en as a fact and, because "reports" need to be confirmed before becoming "facts,"
a particula r actIo ' . e caller (B) ha s Jus t gIV
(Whalen et al., 1988), the sked follown. � g ep1��de;c�urs�Thquest ion design, whether
she also avoids committing the school to an "official" position on the issue.
At the same time as professionals and institutional representatives are
oic "
his address an d th en is a
or an
, u
apa
� I�� a : x
1 �
en s he data show, he reshe ponds with often cautious about making claims, they also deploy distinctive, functionally
h ouse
the ad dress is a
ctio n draw n not fro m t h � � ' s he is given, but rat r from the
hOlce specialized, and superior knowledge bases that can impart a specific expert
a lexical sele . , It IS a home" ..
, authority to claims made within the relevant knowledge domain. The episte­
e of real e state . He rep lIes: mological superiority of expert knowledge is sometliing that is recurrently
languag
renewed in talk and in many different ways (Silverman, 1987; Gill, 1998; Jacoby
use or n' apartm en'?
A: Okay iz th is uh ho and Gonzales, 1991; Perakyla, 1998, 2002; Raymond, 2000). Medicine provides
B : It- it Is a ho: me numerous examples. Patients may orient to the authority of medical know­
der _
the press ure of the emergency (the caller 's mother is ledge by their lexical choices, for example, the tentative or uncertain use of
Here, prob ably u �
ply . s to gras the relevance of the distin on between
fall
cti medical terminology (Silverman, 1 987; Drew, 1991; Maynard, 1991), or by
SIm ss and
dying
a hou
),
se
the
an d
calle
an
r
a part m ent to an am b uI ance crew looking for an addre failing to raise questions about important problems and concerns (Frankel,
1990; Todd, 1993), or by permitting "medical" definitions of their problems
a way to enter th� lo�aOnal tion. . d by one
orga mzatI c ontIn genCle ' s which are taken for grante kinds of
to prevail over their lifeworld concerns (Mishler, 1984). Moreover, lack of
Ro utine
but a re un kn ow n to the oth er, can be the source of many W
ot. her
ha 1 en (1995)
medical knowledge may mean that patients may not know or understand
the purposes lying behind particular questions, and they may not grasp the
party
nfU . n. In the c�se of "91 1" emergency calls,
SIO
difficulty a nd co of the cu rso r on line of inquiry which the doctor is pursuing in questions on what seem to
argu e d that such conti nge ncIeS a.s the current position wh ich qu est ion s be unconnected topics. This lack of access to the "hidden agenda" of doctors'
has
nu-d n . ven computer screen can 111f1uence the order in nt to ca 11 ers . questioning represents another avenue of analysis into asymmetry in medical
a me t h em seem confusing or irreleowhow" often va
are a sked , an d so e
� . tim es ma ke
. ational and interactional "kn interaction (Fisher, 1983; Silverman, 1987).
amz . ns
Similar asymmetnes In org . ' on s , an d oth er int era ctlO
strongly influence porIcereso and courtroom 111 t err og atl
urce;;�d routines are used to1;eva luate the truth Rights of access to know/edge
orga niz atio nal 1990). Asymmetry of knowledge arises when people - usually lay people - have
in whi ch
clai ms (Bo yd, 1998; Drew , 2 '' Heritage et al., 200 Watson, limited resources with which to answer the questions "what do I know?" and
of la y
"how do I know it?" But these same people may also have limited resources
knowledge with which to answer the questions "what am I entitled to know?" and "how
and asymmetries of
. al cau tIOn
.
cti on is a kI. �d. of
Epistem ologIc
ny kin ds of ins titu tio na l int era am I entitled to know it?" Limitation in this regard is an asymmetry in rights
A notable feature of masne ,, In the professionals avoid comm lttmg
. . O��� This cautiousness is mandatory In tat' n
' l '
llC h .
epistemo1 oglca I " cauti ou ss w cer of access to knowledge. Here lay persons are sometimes in a position ana­
. flIm. POSlh
themse�ves �
logous to the gossips described by Bergmann (1993): they have information
.

t t aking . such a erview (Heritage, 1985; Claym an,

t ract lOns new s int � Dr e , that is relevant or significant, but they do not have rights to know it or they
institutlOnal :
1988, 1992; He
In
ntag e and Gr � at � a t c h 199 1) or the courts (Atkinson a� IS qUl�e
dical diagnosis, It . :15
have come to know it in a "morally contaminated" way. Thus a person calling
2). In ot e� c�nt�xts such asofmeepis to inform the emergency services about an incident may be at pains to show
1979; Atkinson, 199 . Ms Matahn s ca11s 'a k1'nd
In
temological cautlOn
away that they are calling from a sense of duty about an event that imposed itself
common. E
or
ve n
exa m P Ie
·
wh en a pa ren t see ms unaware that their ch. 1'ldthIS'IS' way'. on them, and not because they are "nosey" or "looking for trouble" (Whalen
evident. F . norma11y te11s them about the absence 111
from sch ool, Ms. Mat 'al111 and Zimmerman, 1990). A patient who is concerned about a possible illness
may be similarly at pains to show that he or she is not excessively preoccupied
Sch : Was Willia
. , ill tod ay?
m h ome from sch ool - .
ano the r person
with minor bodily changes (Halkowski, 2004). Patients are similarly reluctant
off the p h one In whic h Mo m asks to voic e diagnostic hunches about their illnesses except under relatively
2 . . . « conversation

3 . . . if Wil liam was

home)) defined circumstances (Heath, 1992; Perakyla, 1998; Gill, 1998; Stivers, 2002;
4 Mo m: No h e wasn't
239

238
JOHN HERITAGE TALK
CON VERSATION ANA
LYSIS AND I N STITUTIONA
L TALK
Heritage and Robinson, 2004), and Strong (979) documents the fact that CONC L U S I O N
doctors accompanying their children on pediatric consultations suspend
their medical expertise and act "like parents" when dealing with the attending B� �o�, read ers will have seen
that man of the . ' rent dIm .
physician. In this last case, persons with every "right" to medical expertise of /Instltutionality" in talk are �Interr dIffe ensions or els
voluntarily suspend those rights in the limited environment of a medical that fit insid e one another, each 0 th orou g hI y elat�d. Rather like l�ussianlev dolls
consultation with another person qualified as expert. In institutional inter­ 1eveI : IeXl. cal choice is part of turn f these elements' IS a part of. the
nex t high er
action then, knowledge may not be enough; one must also be entitled to the organization; sequence organizaf10 'd eSl' gn,' iurn d eSlg
, a
c

' n IS. a p art of sequence

knowledge, and have come to it in an appropriate way. There are two "wild card s'" In thn IS a Part of overa.ll structural organiza tion.
a d·.ISt'Inct turn-takI. ng system IS" In e pack. Turn takI '
ng IS' one because where
an Interaction's organiz"atl'on . Asy P J ace' It' has m aJor effects
- mmetry IS the other beca usate man y levels of
.

F R O M Q UALITATIVE TO RESEARCH: at all other levels of the organ'lza . , it is em bod ied

A N OTE ' choice turn d esl'g n, seq t'IOn 0f Inte
1 eXlcal ' ractIOn " In " InstItutI onal sett ings'
" tlOn, and tur
lza 'n uence organizat'lOn, overall str
taking. . Indeed CA may end ucture organ-
As many readers will have inferred, if the kinds of qualitative observations Foucauldian conception of po up WI'th - ap . nity with a rather
made in this chapter are true, then they should have an impact and significance wer, a dvoca ted by �ther affi .
. , .
vo1ume. The view that powe contnb utor this
that is quantitatively measurable as well. For example, questions that are fications, kn owhow, and norm r inh er es In InShtu ho?al knowledges, to
CA view that it is crea ted renewative a�angem�nts �s comp atible with the
designed to favor a "no" response, e , g., negatively polarized questions such c1assi­
as "Any questions?" (Heritage, 2002; Boyd and Heritage, 2004), should result but interlocking facets of 'the orgam ed � an oper tlOnahzed in man
, � y dispar
in less questioning overall than positively polarized questions, e.g., liDo you .
converge In the idea that thO ' zat
. lOn of Int erac,t'IOn. Both perspectivate es
have questions you would like to ask?" Most of the work that has explored ficatory, and interactional pr:cE�ewSer �nhe,res .both in the know ledge, c1assi-
, the d'Iscre . O InSt1 tutlOns and t1lelr. Incum
quantitative aspects of institutional interaction has focused on this level of In tlOnary freedoms wh'ICh thfose practices permit for the incbent
.
s and
turn design, For example, Boyd (1998) distinguished between "bureaucratic" of institutional roles. umbents
and "collegial" opening questions in interactions in which physicians were
questioned about patients' need for surgery and found a systematic relation­
ship between the "collegial" opening questions and the likelihood that the
patient would be approved for surgery. This research can also have an histor­
ical dimension: for example, Clayman and Heritage (2002a, 2002b) showed
that question designs that are more hostile and assertive have become very
much more frequent in presidential press conferences over the past fifty years ,
This quantitative work can have significant "applied" value. Stivers (2002)
distinguished between medical problem presentations that simply describe
symptoms ("sore throat, scratchy cough, phlegm") and those that suggest a
diagnosis ("strep throat"), and found that physicians were more likely to
believe that the latter problem presentations reflected a desire for antibiotics
(and prescribed accordingly) even though the differences did not reflect actual
patient desires as revealed in pre-visit surveys (Stivers et al., 2003). Similarly
conversation analytic research that suggested that IIonline commentary" (e.g.,
"Your throat's a little red") produced during patients' physical examination
might be a means of forecasting "no problem" diagnostic outcomes and
avoiding unnecessary prescribing also turned out to have quantitative support
(Heritage and Stivers, 1999; Mangione-Smith et al., 2003). RE FE R E N C E
These are relatively straightforward linkages that focus centrally on turn S
design - the easiest dimension of interaction to quantify. They suggest that 6 )
. (19 4 . "Rhetoric " "lOgl'C "
the study of institutional talk can support an important IIappliedII dimension
Albert
' " I' n B r nd '
E
" i t 66 pt (6) 3
2 'u u
and "
-54
t1CS
in the future, though considerable work will be needed to achieve this 1 ehavior. Americ"an Anth ropolo o- s poe
SPeecI b
1: cul tu re patterning of
Atk mson (
J M, 1 982) . Und erstan , .
] Iity.. notes on the
' S

objective. , .' . ding forma ,

' I7 ournal of
prod ch
6
u on of "forma l " interacti categorisati on and
on . B rz'tIS
Sociology 33: 8 -1 1 7.
240

241
 CONVERSATION ANALYSIS AND INSTITUTIONAL TALK
JOH N HER ITAG E \\ TALK

Atkinson, J.M. (1992). DiS
roceedings. In P. Dre199-
bniversity Press,Drepp.w, P. ( ��;9) , Order in Court: The Organis
'
Atkmson, J . M . and.
in Judicial Settmgs. Lo�d n '
Atkinson, J.M . and H �s.ntag
. . . formal aspects of informal c?urt organization of mediation hearings minimizes argumentative talk. American
w an .
P�r��ri�:;:����', Talk at Work. Cambridge: Cambndge
Sociological Review 56: 818-35.
.
Gill, Virginia (1998). Doing attributions in medical interaction: patients' explanations
. ation of Verbal InteractIOn
for illness and doctors' responses. Social
Goffman, Erving (1955). On face work PsychPsycho logy Quarterly 61 (4): 342-60.
iatry 18: 213-3 1.
' Macmillan. . Goffman, Erving (1983). The Interaction Order. American Sociological Review 48: 1-17.
e , J .. (eds. .) (1984). Structures . ofy Pres ' . Studies in
SOCia I AC tIOn.
Goodwin, Charles and Goodwin, Marjorie Harness (1997). Seeing as a Situated Activity:
0

Conversai:ion Ana. lysI cambn

dge . Cambrid e UniversIt clients:s. a study of verbal Formulating Planes. In David Middleton and Yrjo Engestrom (eds.), Cognition and
Baldock, J. and I)nor. , D . nal (1981) . sOCla I w orke�s talking) to comm unication at work. Cambridge: Cambridge
rk 11: 19-3 8. Goodwin, Marjorie Harness (1996 ). InformingsUnive rsity Press.
.
behaviour.. British Jour3). Disc of Soci al WO
. . ' .
Ip and Announcements in Their
t ' ns ' The Social Organiza tIOn of GOSS .
Bergmann, Jorg. R . (199 Environment: Prosody within a Multi-Activity Work Settin
. In d.lscre
reet
Kuhlen and Margret Selling (eds.), Prosody in Conversationg.: InIntera Elizabeth Couper­
10 .

Chicago: Aldme. eaucratl..c auth on'ty 'in theAme"com pany of equaIs . the
B oyd, Eliz. abeth (199 8). Bur Cambridge: Cambridge University Press, pp. 436-6 1. ctional Studies.
Sociological Remew 63
ft .

interactional management of medical peer revlew. rican . HIst . ory..

Greatbatch, David (1988). A Turn-Taking System for British News Interviews. Language
(2): 200-24. in Society 17 (3): 401-30.
Heritage , Joh (2 4 Tak in the Pati ent' s MedIcal Halkowski, Timothy (2004). Realizing the illness: patients' narratives of symptom
Boyd, Eliz abeth and � �s��ry Takfng. In John Heritage and Douglas
Quest.10' m. ng During Com prehensIve
. , g M ed ICI " �e.' Structure and Process . Primary Care Encounters.
. discovery. In John Heritage and Douglas Maynard (eds.), Practicing Medicine:
Maynard (eds.), Pracidge tlc .
m
Structure and Process in Primary Care Consu
ltation s. Cambridge: Cambridge
Cambridge: Cambr U �����\ti97P���.s[)octors Talking to Patie� ts : A St�� f �:
.
University Press.
m

Byrne, p.s . and Long, B.E. L. al College e� e � Heath, Christian (1992) . The delivery and reception of diagnosis and assessment in
. the Cons ultat ion. Exet er: Roy the general practice consultation. In P. Drew and J. Heritage (eds.), Talk at Work.
0
urs of Doc tors
Ver bal
. . ners .
Beh avio m

Practitio . view . s . Social Problems Cambridge: Cambridge University Press,

. .
Clayman, S. (198 8). Displaymg neutrality in television new
s mter Heath, Christian and Luff, Paul (2000). Technopp.logy235-6 7.
in Action . Cambridge: Cambridge
35 (4): 474-92. . m' th C . v ent of neutrality: the case of news mte . r�lew . University Press.
Footmg Heritage, John (1984). Garfinkel and Ethnomethodology. Cambridge: Polity Press.
and J. �e��:g:7edS.), Talk at Work. Cambridge: Cambndge
Clay man, S. (1992).
discourse. In P. Drew163- . s
Heritage, John (1985). Analyzing News Interviews: Aspects of the Production of Talk
University Press, �p. 9�002a). The News Interview: Journalists and Public. FIgure for an Overhearing Audience. In Teun A. Dijk (ed.), Handbook of Discourse Analysis
CIa:�anAS � and Hen tage , J. ( . ve s ' Pr Volume 3. New York: Academic Press,
he i . Cambridge: Cambridge um � �r ��: nalistic deference Heritage, John (1987 ). Ethnomethodology.pp. In
95-119 .
Antho
Clayman' S. and Heritagm. e,thJ.e (200 2b). Quest ng sidents: Jour Reagan. Journal of (eds.) Social Theory Today. Cambridge: Polity Press,nypp. Giddens and Jonathan Turner
and adversarialness 749-p5.ress conferences o f Eise nho wer and Heritage, John (2002). Ad hoc inquiries: two preferences 224-7 2.
0

. .� in the design of "routine"

Communication 52 (4): ?es 0f knowledge in conversational interact1Ons questions in an open context. In D. Maynard, H. Houtkoop-Steenst ra, N.K. Schaeffer,
P . (199 1). Asy mm etrI . and H. van der Zouwen (eds.), Standardization and Tacit Knowledge:
D rew,
I . Markova and K. Fopp a (eds.), Asymmetnes In, Dialogue. Hemel Hempstea . Practice in the Survey Interview. New York, Wiley
Inters cience , pp.
Interaction and
313-3 3.
Harvester Wheatsheaf, ppeV. Idence
tested .29-48. . n a c urtroom cross-examm. at10' n '. the case . f a Heritage, John and Greatbatch, David (1991) . On the Institutional Chara cter of
Dr w P (199 2). Con : t . at Work. Cambridge: Cambndge Institu tional Talk: The Case of News Interviews. In Dierdre Boden and Don
a

�ri�l f�rrape. InPs,. pp. DrewandJ. Hentage eds.) , Talk Zimmerman (eds.), Talk and Social Structure. Berkeley, CA: University of California H.
University Pres 470-520. k: An Intr . . In
oduc.tIon Press, pp. 93-137.
Drew, Paul and Herita ge, John 1 99
� �\t7 J a I zin g Talk at Wor Heritage, John and Lindstrom, Anna (1998 ). Motherhood, medicine and moral
. : Cambridge U mversity
P . Drew and J. HerItage (eds.) , a a ork. cambridge scenes from a medical encounter. Research on Language and Social Interaction 31 (3/4): ity:
Press, pp. 3-65 . ics. Berk e1ey, CA. UniversIty
. of 397-438.
Duranti, Alesandro ( 1 994) . From Gram mar to Polit Heritage, John and Robinson, Jeffrey (2004). Accounting for the visit: givin
. s are negoti. ated for seeking medical care. In John Heritage and Douglas Maynard (eds.), Practi g reasons
California Press.
Doc tor talk < pat . ent �alk h treatment decI. sIon
�
Fisher, Sue (198 3). A. Todd (eds .), The 50t� . I
cing
mcat1On. 5 · �:her and ton, Consultations . Cambridge: Cambridge
Medicine: Structure and Process in Primary Care
in doctor/ patient commu mun�lcat ' n . Washing DC: Center for e App I University Press.
Heritage, John and Sefi, Sue (1992). Dilemmas of Advice: Aspects of the
Organization of Doctor-PatIent Com Delivery and
10

Lin uistics, pp. 135-57. mg , rvie. ws . a dispreference for patient initi. ated Reception of Advice in Interactions Between Health Visitors and First Time
Frank�, Richard (199. .0).n-pa Talk . m. mte ' In P. Drew and J. Heritage
(eds.), Talk at Work. Cambridge: Cambridge Unive Mothers.
ent encounters.. In G . Psathas (ed.), Interaction Compete rsity
nce.
uestions in physicia fIPres A�enc�, p. 231-62. how the interactl. OnaI
Press, pp. 359-417.
Heritage, John and Sorjonen, Marja-Leena (1994). Constitutin
�anh am, MD: University s oflut1O
GarCIa, Ange1a (199 1). Dispute reso n WIt�ou t disputin
. g: g and maintaining
243
242
 CONVE RSATI ON ANALY SIS AND INSTIT UTIONA
L TALK
JOH N HER ITAG E " TAL K

. gn. Language Sacks, Harvey (1979 ). Hotrodder: A Revolutionary Categ

-pre facmg as a feature of question desi ory. In George Psathas (ed.),
activities across sequences .' and Everyday Language: Studies in Ethnomethodo[ogy.
. ' 't .
New York:
'

Commentary in Acute Medlca1 VlSI s.

c

Irvington, pp. 7-14.

in Society 23: 1 -29. Sacks, Harvey (1987 [1973 ]). On the Preferences for
1-�
Agreement and Contiguity
' I Science and Medicine 49 (11) : 1 5� 7.
Stiv ers, ran ya 099 9 ) . Onl ine
Hen' tage' John and in Sequences in Conversation. In Graham Button and John
c

. RE. Lee (eds.), Talk and

, hod, of Shaping Patient Expectat1?ns.
SoCla
A Met (200 1). Sub vert ing Cnt ena: Social Organ isation. Clevedon, England: Multilingua
l Matters, pp. 54-69.
and Klemman, Lawren ce .
Heritage, John, Boyd Eliz. abeth,. ery. y af Heal th and lllne
ss Sacks, Harvey (1992 [1964-72]). Lectures on Conversation (2 Vols.)
. Oxford: Basil BlackwelL
The role of precedent m d eClS10 " ns to finance surg SocI Olog
Sacks, Harvey, Schegloff, Emanuel A., and Jefferson, Gail (1974)
. A Simplest Systematics
, 0

. .
23 (5): 701 -28 .
1 atn '
. k (199 1). The Conshtu
c
. tlOll
' 0 f Expert-Novlce m
for the Organization of Turn-Taking for Conversation.
Language 50: 696-735.
d Gon zale s I
' Schegloff, Emanuel A. (1982) . Discourse as an interactiona
Jacoby, Sal y an J
Scientific Discourse. Issues In A�
. p l'Ied Linguistics 2 (2): 1 49-81 .
on as an Interactional Resource. Lang
uage In Society. of "uh huh" and other things that come between senten
l achievement: some uses

Jefferson, Ga1' 1 (197 4). Error Correch

ces. In D. Tqnnen (ed.),
0 '

Analyzing Discourse (Georgetown University

Roundtable on Languages and
e y namics. In 1. Markova and 1<. Fop
Linguistics 1981). Washington, DC: Georgetown University
2: 1 81-9 9.
Linell, P. (199 0). The power of dla
. l �� ; Wh eats hea f, pp.
pa (eds .),
1 47-77. 1 Schegloff, Emanuel A. (1984). On Some Questions and Ambi
Press, pp. 71-93 .
guities in Conversation.
The Dynamics of DIalogue. cHern
e e stead: Harvester
1 ell' P. . and Luckmann, r . (199
�
1). A Y mmetrie s in dial ogu e: som e con ceptua In J. Maxwell Atkinson and John Heritage (eds.) , Struct
Cambridge: Cambridge University Press, pp . 28-52. '
ures of Social Action.
. man' es. . In I . Markova and 1<. Foppa (eds .) , Asymmetries in DIalogue
L'n . Hemel
Schegloff, Emanuel A. (1986) . The Routine as Achievement.
. ..
P re hm , f, pp. 1 -20.
,
Human Studies 9: 111-5 1 .
.
Hempstead·. J-Iarvester Wheatshea P. (1988). Interact'lOna1 dominance In dyadIC Schegloff, Emanuel A . (1992 ). Repair after next turn: the
last structurally provided
P., Gu stav sson , L., and Juv or :en' . . . defense of intersubjectivity in conversation. American Journa
LineH, . .. s. LIngUIstIcs 26. 41.5-42.
tat1On of Imhative-response analysi
l of Sociology 95 (5):
commum' catl'on'' a presen tage,1
' Hen 1 295-345.
. . , ya, E lliott, Marc, M cDonaId, Laurie and. Schegloff, Emanuel A. and Sacks, Harvey (1973) . Opening
. On too
ManglOne-Sml' th" Rita StlVers, Tan
or1 phY sical exam f''md mg mum cat1 ' Up Closings. ScmioUca 8:
�

' s '. a com . . 289-327,

John (2003). Online commentary . . cn 'b ' ?. Social Science and MedICIne 56 (2).
mg
.
.rd111
' g inap pro pria te anh b10t lc pres Silverman, D. (1987) . Comm unication and Medical Practice. Londo
for avo n: Sage.
. teractional and institutional bases 0 f asymmetry 111 c1"mIca, 1
. Stivers, Tanya (2002). "Symptoms only" and "Candidate
313-20. diagnoses": Presenting the
Maynard, D. (199 1 ). On
·
dIScou
t h e
rse . American Journal of SOCIO Iogy 92
111
. (2): 448-95 .
. everyd ay 1'£ " " ,
problem in pediatric encounters. Health Communication 14
Stivers, Tanya and Heritage, John (2001) . Breaking the
(3): 299-338.
sequential mold: answering
1 e. American SOCIOlogIcal ReVllW
Maynard, Doug1as (1996) . On "rea
liza tion " "more than the question" during medical history taking.
Text 21 (1 /2): 151-85.
111

Stivers, Tanya, Mangione-Smith, Rita, Elliott, Marc, McDo

60 (1): 109-32. Classroom. nald, Laurie, and Heritage,
1 of Turns at Formal Talk in the
McHou1 , A . ( 978). The Orgamz
at'on John (2003). What Leads Physicians to Believe that Paren
ts Expect Antibiotics? A
'
c

n A. D IJ"1( (ed .),

213 . Study of Parent Communication Behaviors and Physicians'
Lan uage in Society 7: 1 83- Perceptions . Journal of
Meha Hu ;' gh (198 5). The Stru cture of a room Discourse. In Teu
S volU ;;; � ew York: Aca dem ic Pre ss, pp. 1 20-31 .
Family Practice 52 (2): 140-8.
Strong, P. (1979). The Ceremonial Order of the Clinic. Londo
Handbook of Discourse AnalYSI . '. s of Med ical Inte rvie ws. Norwood, Todd, Alexandra (J 993). Exploring women' s experience
n: Routledge.
. hIer, E . (198 4). The Discourse of Me IClne. Dialectic s: powe
MIS medical discourse. In Sue Fisher and Alexandra Todd (eds.), r and resistance in
I'
NJ: Ablex. . . I PractIce . . The Social Organization
ssl ) . A IDS Cou nsel ling ; Institutional Interaction and C I11lCa of Doctor-Patient Comm unication (2nd Edition). Norw
ood, NJ: Ablex, pp_ 267-86.
Perakyla, An (199 5
. ' Press Watson, D.R (1990) . Some features of the elicitation
Cambridge: Cambridge Umverslty . . of confessions in murder
and acco un t a b1. lity: the delivery of diagnosis 111 pnmary interrogations. In G. Psathas (ed.), Interactional Competence
Perak la, A. (199 8). Authon ty
r rt rl� 61 (4) ' 301-2 0. . Press of America, pp. 263-96.
. Lanham, MD: UniverSity
he lth care. Social Psychology Qua �
author ty � ci d responses to diagnostic stateme ;.
nts Whalen, Jack ( 995). A technology of order production:
Perakyla, A. (2002). Agency and
in primary care encounters. R�sear
c ����; � ua e and Social Inte ract ion 3 � (2): 219 47
; public safety communications. In Paul ten Have and Georg
computer-aided dispatch in
e Psathas (eds.) , Situated
Pomerantz, Anita (198 8). Off enn g a � an l
d'd ate Ans wer : An Info rma t1On See lUo Order: Studies in the Social Organization of Talk
DC: University Press of America, pp. 1 87-230.
and Embodied Activities. Washington,
hs 5 . 36�_73. . of
Strategy. CommunicatIOn Monogra�
Raymond, Ge offrey (2000 . �he vo r � f ority: the local accomplIshment
: hve news broadcasts'. Discourse Studies 2(3): 354-:9 . he
ce auth Whalen, J. and Zimmerman, D.H. (1998). Observation
of emotions in naturally occurring activities: the case
s on the display and management
authoritative discourse 111 of "hysteria" in calls to 9-1-1.
. urces m t
mmetry m actl'on'. sequential reso
Social Psychology Quarterly 61 (2): 141-5 9.
ins on, Jeff rey D. ( 2 001) Asy Whalen, Marilyn and Zimmerman, Don
Rob . .' uest · Text 21 0 /2): 19-54. . H. (1990 ). Describing troub le: practical
negotiation of a prescnpt10n req . 1 Structure of Med ical Act
iVIt . Ies. DUrmg epistemology in citizen calls to the police. Language
.
D . (200 3) An Inte rac t lOna Whalen, Jack, Zimmerman, Don H, and in Society 19: 465-92.
Robinson, Jeffrey . mUrllca t10n
Acute Visits and Its Imp]'lcaf10ns
for Patients' Participation. Hea lth Cam
A Single Case Analysis. Social
Whalen, Marilyn R (1988) . When Words Fail:
15 (1): 27-59. . .
actIVlty transltlonS 1
. . 'n
Zola, Irving K. ( 987). Struct
-
Problems 35 (4): 335 62.

Hu 11
.
Robinson, Jeffrey D. an s J � d t rs � :��!t� Ach ievi ng
ry taki�g to physicial examination. ma
(200 1)
of non-compliance. In Howa
ural constraints in the doctor-patient relationship: the case
rd Schwartz (ed.), Domin ant Issues in Medical Sociology.
rimary-care consultat1Ons. ro New York: Random House, pp. 203-9.
�omm unication Research 27 (2): 253--98
.

245
244
 THE CONCEPTUALIZATION AND ANALYSIS OF VISUAL DATA

a picture of an adult and child of indeterminate sex, both dressed identically

in jodhpurs, riding jackets and hats, astride white horses on a deserted beach.
Underneath, offset to the left, a balding man in his forties dressed in casual
attire is shown scooping the leaves from the surface of a backyard swimming
pool. Below this a third photograph depicts a smiling bikini clad woman
in her late twenties. Cradled in her left arm is a toddler whilst her right
arm is hooked round a fallen tree stump. The pair are seated in a shallow
free-flowing mountain stream. A single, three word, caption accompanies
the images: 'Water and Class'.
What is going on here? What are these images trying to tell us? Equally
importantly, how do they tacitly trade upon the viewers' commonsense
zatio n a nd
c o n ce ptu ali reasoning to achieve this goal? We quickly learn that the photographs are
part of a 'visual essay' designed to accompany Chapter 11 of the book The
of vis u a l Architecture of Disadvantage: Poverty and Wealth. Similar" visual essays' can be
found at the end of the majority of each of the chapters. Somehow, then, the
on images are purporting to present us with scenes and information relevant
Mic hael Em mis
to these sociological concepts.
Visual researchers frequently allude to the popular maxim that their photo­
graphs 'speak for themselves'. But in 'Water and Class' - and many others like
this - it is not the photographs which generate whatever sociological insight
is claimed on their behalf, but the viewer. That is, it is we who make the images
speak, we who do the work of 'class analysis' not the photographs . One key
. g WI'th 'visual research and
I NT R O D U C
T I ON , component here is our commonsense understanding of the conventions
ks deah of textual layout and what they signify. That is, we make the assumption that
f he existing textboo
Pick up almost any 0 t e discussion will focus �n .som
� e :eaJ on the use of
� � :
��� � � ranch of visu al the first photograph is a depiction of an 'upper class' scene on account of its
location at the top of the page; conversely the representation at the bottom
it is a safe
� \: �
photo graph � mm in fr om
cal ly ov e im pl
: es. Dramatithe Anglo-saxon t � . of ethnogra
rs
th e g en era
ph y and
tion of a must be that of the 'working' or 'lower' class. Moreover, we provide additional
research, that steogy, a�vocates the use of the cameranm �l'th closer affinities narrative material locatable in the images which serves to reinforce this inter­
social anthropol e r earch setting; a second branch, od � S favours the inves­
�
visual reco:d of t sc��ols of semiotics and cul tural StUnal� SI' . s of their implicit
pretation: only the upper class can afford the ownership of horses; the deserted
beach must be part of the riders' private estate and so on.
to the contmenta . 11 roduced ima ges and an a Y 'the visual' with In making sense of these images in the way I have suggested, something
tigation . of co::C�l��r J �essages. E quatin g the :���Ic: f it, so self-evident
; very similar to what Schegloff (1988), in another context, has referred to as
l 'sociology by epitome' can be detected. Schegloff's target in this instance
: ::�����o�ectio� orgsa;:�i�i����:;;�:::���e constr ued:C�f�:����
h o t°
pe e : was Erving Goffman and the widely held view that his (Goffman's) work
that to suggest .thm is is pr ecis ely w ha t I .
e. And yet th an argument for SOClal researchers cal and int ereste d was densely empirical. In Schegloff's view this was misplaced; Goffman's
even p rovocatwt fo nt piri observations of social life, he argued,
h llows I mou
�ha��;��:g �isual material in their wor���/��:ain. In summ ary from
h ' k both the em
achieve their sense of typicality, by using but a stroke or two, . . . a detail or two, to
C
:n:lytical poss�ISbilt�i:�i: ��ht�� �:J��� terms of nyott0o�l�g��t �ack
: ! indicate the scene which we as readers are to call up from memory, personal
.
e eye ca� see. 0nefusef
experience or imagination. If he succeeds, that is if we succeed in calling such
case I develop what th · ul wa
l enqUlry.
concrete mstance 0
vis ua a scene to mind, our ability to do so from his detail or two is proof of its typicality.
can re cor d b ut i mag e-b a se d
. The typicality of the scene or action has not only been 'shown' but has been
claim is to examme a enlisted and exploited, and the adequacy of his description . . . has ipso facto been
demonstrated. (Schegloff, 1 988: 101)
?
F O R IT S E LF
TU R E S P EAK
'te (Newman
Much of conventional image-based visual enquiry operates, I argue, accord­
ay l 1)'
chitecture of everyd
P I C
A ,
.
DO ES
ing to the same method of epitome. The photographs which provide the
S oci ology: explorm h to the
Pag
.
e 38 3. o f
ree black an� tw ��;e photographs. At the top,
IS occUpled by th 247
iiIi .

THE CONCEPTUALIZATION AND ANALYSIS O F VISUAL DATA

VISUAL DA TA
MIC HAEL EM MIS ON
6

rts and journals2 what one contributor has referred to as a 'deluge' (Pink, 200lb)
. ld's ethnographic essays, research repo
-

for th e fle supply the which have recently made their appearance, testifies to the legitimacy, if
illustrative m aterial a purp se on1y to the extent that we cand Class' m y not unqualified acceptance, of the field in the eyes of social and cultural
m ono graphs serveptual po?mt th.ey purP ort to deliver. 'Water an ess the pomt � researchers.
ce pr
ple of th�� f�nre b t of
theor etica l or c on it is not aty pic al. To
e has been At the same time, however, and partly as a consequence of this increasing
be an extreme examnt, 1 e o r d �r the images on the pag wer? Most interest, questions remain about the viability of visual research as a coherent
further for a mome is have led to conf�sl�n on the part of the the
w hat . vie
th ve capacity intellectual endeavour. The multi-disciplinary nature of the field is such that
reversed? W ould nkel (1967) fa o owed huma ns ha s have been
likely not. As Garfipurpose even m s��J:g: whe;e these feature bottom of
� it is no longer possible -- if it ever were - to speak of th.e method of visual
enquiry, not least because there is no obvious agreement about what this
to find sense and se the two horse l. d S noW relegated to theclass jockeys term should embrace. Significant fault lines can be detected on the basis of
violated. In this caell have bee cons�rU:� as apprentice working bikini-clad even a cursory inspection of the literature. For example, as we have noted
the page - might w e ClSm . . ' g the�n charges at dawn. Conversely the a weaIthy above, there is no consensus as to whether visual researchers should generate
or strappers out py�mg the top billing, could well have been seen as
ex
eat. It would the material they seek to analyse or alternatively conf:i.ne their attention to the
female, noW occu in the clea aters f her private mountain retr
heiress luxuriating y the b a\:m . g, c�sually dressed man as
_

anyt hin g other analysis of pre-existing cultural images and representations. Both practices
de ntif are widely advocated but both draw upon divergent theoretical traditions and
be difficult to i semy' to u�e demand correspondingly different analytic stances. Tensions dating back to
than middle class.] very roblem of indeterminacy, or 'polypho
� I h:�e identified in the use to graphIc the realist and artistic traditions from the foundation period of photography
Paradoxicall . y,
J
the
arg W h �lC -
cultural studies rned to a van age by some researchers. taryson
on,
of
As Ma (2002) have resurfaced in new guises. There are those who insist that visual images
are inherently interpretive: a domain of genres, narratives and codes. For
ma terial
images have bee phs' and other forms 0f VISU
n tu ' al or docum en
n . other
wIth others photographic information is a faithful record of what occurred and
ogra
has noted, phot met od 0f data generation in conju nctio
'pho.to­ which might even be quantified. Traditional and experimental ethnographers
have b een use d asntervI�ew�mg or focus g-rouP s. In the methodintoferpretmg alike who have added video to their repertoire of research techniques ply their
metho ds such as i s upon the tas k of
elicitation' (Schwartz, 1989) the. resear1 cher1 t ademu
l'

nity to generate extenslVe

.
trade in apparent ignorance of the use of the video record in the ethnomethod­
existing photog rap hs
y w IC
� �
� m
m
� l
g
Y
h � � o �� e � �:e rwis e forthcomir:g durinphoto­
theIr own
g the ologically influenced analysis of naturally occurring forms of interaction
(see Heath, this volume). It is not putting too fine a point on it to state that the
verbal comm entarAlt ern ativ ely, su b ' cts can be asked to takesion
Je and analysis field of visual research stands as one of the most disorganized and theoretically
interview pro ces s.
bas �
. of subse uent discus inchoate in the social science academy.
graphS which canGe then be th e
0 � otog raphy' , this method has
by the researcher. stiner
ally refe rre
phe n
t
o
as
me
'au
na
t
a f h
sex role differences (Claillncy
dentity (Z er,
The principal message I want to convey in this chapter is the need to think
of visual enquiry as embracing more than the photographic image. There are
f1ons of n ation al i
gate suc h
been used to inve or cep
and Dollinger, 1 993)
cro ss -cu ltur a l per several strands in this argument which require unravelling. The first, and
less contentious, is to think of images not simply as a realm of representation
1990). but also as containing information or data which can be bought to bear on the
M I NAR Y investigation of social and cultural processes. It is this theme - that the visual
SO M E P R EL I i� a realm of 'data', not simply a domain amenable to 'cultural' or 'interpre­
tIve' modes of inquiry - which is intended to be conveyed by the chapter
. s the case that until. rela.tively title. In part, thinking of the visual as data may require going beyond the
. . . no�w� 'th st andi ng i t rem am sCl:nces reliance on the photograph and considering the possibilities inherent in other
These possiblhtles maJ q a l itati ve res ea rchers in the soclal log
lty of
recently, the vast m alo�mterest :�n VISU . :1 enquiry or visual methodo Ies. Ina forms of visual material of the kind I shall refer to later as two-dimensional .
re has been For example, newspaper cartoons or comic strips can tell us a good deal
h ave had only minihowever, thIS h;S eg�.n to change and themu ation, in about the wider political, economic and gender systems in which they are
burgeoning interes � :
the last few years, t in all aspect 0 , he �:sual' in visual com nic
ur e m ore broadly. 'The visuant
_
al' has embedded. Other forms of two-dimensional visual data such as directional
atio n and Vis ua cu nds of hro- signs, diagrams and maps can be used to explicate the claims of ethnomethod­
Practices of visu1yaliz s of c on ce rn . I'tS traditional homela
m . gaged
. ology about the role of commonsense reasoning. Here the focus is not so
b ecom e not on a an focu so m ething WhICh has endes
.

st die s b ut
polo gy, sociology drs �m Iscc ltu al
: . � a� diverse as architecture and lg�, much on the discovery of cultural meanings by the academic analyst but
the interests of schola
nd urb an stu
�
dIes,
�
��:�al culture, new technology abondokmu1tl
of n ew s an
d :ather the ways in which ordinary actors use or make sense of such visual
Information in the course of their everyday practical routines.
geogr. aphy a ora
museum stud'les. The pleth
medla, and science and 249

248
 THE CONCEPTUALIZATION AND ANALYSIS OF VISUAL DATA
VISUAL DATA
M ICHAEL E MMIS ON
w

al images is Even contemporary luminaries such. as Howard Becker, who pioneered

atin g of 'the vis ual' o nly with such two-dimension is visual in the modern developments of visual So�lO 1ogy in the 1970s, have voiced suspi­
cions as to the appropriatene:' s �)r a equacy of photographic techniques .
But the equ u nduly restrictive. Social lifeue,
sho rt-si ghte d and re are
als o curiou
d
sly
cou nte r-in tuiti ve way s. Consequently, I shall argdverthe tise ment Commenting on the work of OClal d ocumentary ph0tograph ers and their
diverse an othe r a
than the photograph, the rough visual
for ms of vis ual data re1 evance to sociological enquiry, Becker noted a 'double reaction':
many more places and locales carry uage meanings th
or the cartoon . Objects, es. Clo thing and bod y lang are significant signs
At �irst you find that they call attention to a wealth of detm. l from which an interested
means just lik e imag
ty and negotiate public situsoci ations. Eye con­ .
use to est abli sh i denti al life among �oclOlo?ist cou ld develop useful ideas about whose rr:eanll1g he could spin
ll1 te�estll1g speculation . . . . Greater familiaritY leads to a scalll1g down of admiration
which we gla nce' - p lays a role in regulating opp
mel 's 'mu tual ing malls,
tact Sim the built environment - sh ed to exert a
'

mat er ial eco logy of WhIle the photographs do have these VIrtues, they also tend to restrict them�
strangers. Th e been argu
museums and public son paces more generally - has selves to a few reiterated simple sta tements· ' Rhetorically Important' as a strategy
· Ilectually and analytically thin.
th e mov ement a nd mu tual coordination of people. of proof, the repetition leads to work H�a t IS , mte
infl uen ce our uses of
determ ining
bet wee n s urve ill ance , visib ility and privacy regulate ater ial for the
(Becker, 1974: 1 1 )
Tensions s m
there are rich supplies of is only the study
pace s. In all o f thes e area arch
such s up the idea that visual reseran Elsewhere (Emmison and Smith, 2000) I have argued . that It. IS. this reliance on
visual researcher. In giving ge of data b ecomes
'

dver tis eme nts, then a far br oader the photographic image with a11 l' ts attendant problems -.- that has been the
otog rap hs or a enquiry is no longer
_

of ph
ble for inv esti gati on. From this vantage point visualseen and observable.
major impediment to a robust or VI'brant tradition of VISU . ' �1 e�quiry. Visual
availa e , but rather the study of the researchers frequently claim that their field has been margmahzed from the
y of th e imag dimensions
just the stud
aph s may be help ful so metimes in recording the en se core concerns of the social sclences. · I n the case of sociology, where this
Photogr are not necessary . tendency is probably the most pronounced' visual researc 11ers consequently
of social life . Usually they this cha pter I develop these argume nts in more detail. have b een ghettoized and reduced. to communicaf. mg Wlt. h each other about
In the rem ain der of aphically based visual
okin g at som e furth er examples of photogrcogn ized variant. My
a narrow range of specialist issueS My argument IS that this. marginalization
I beg in by lo
as ima ge -bas ed work remains the m ost re ervedly famous is, to a large extent' self-inflicted 'on . account. both 0f the mab Ihty " of visual
rese arch mixe des
d. There have been some al material to make researchers to use photographs m anythmg oth er than a questionable
ict on this ma teri al is
verd have used this form of visu ' fashion but als0 - a pomt
'11ustratlve ' to be devel . ope d m ' more detail shortly
examples of researchers who aps the majority 1

orta nt the oret ical argu ments . However, a great deal - perh nab le. I then turn to to see beyond
. the use of ph o t ograp h y.
imp only be described as questio
_

suc h vis ual en quiry ca n ve an d explore the Notwlthstanding my comments above about the unexplicated comrnon-
_ of
mor e in clus ive con ce ption of visual data outlined abo sense reasoning
. which underpins much p110tographic w�rk, my reservations
the this provides. visual sociolog a1�� turn on the pomt that photographs
possibilities for research which
about an Image-based
J
�ave been misunderstood as constitu ng rn:s of data m their own right when
m fact they should be considered in the f'Irst mstance as means 0f preserving,
M E N TARY AND stormg. or representing information . In thOIS sense photographs should be
AS D ATA: D OCU
P H OTO G RAP H S
R E P R ESE N TATI
O N AL CON SID ERA
T I O NS seen as analogous to code-sheets' th:r�;����es �o mterv .
.iewor
schedules, ethno­
graphic field notes, tape record ing. a1 mteractlOn any one of the
y utilizing pho tographic images ha s produced some numerous ways in which the SOCIa1 researchers seek. ta capture data for
At its best visual en which clearly demonstrate the viability of using visualal
quir
s bsequent analysis and investigation. Ma�' nstream SOCIal researchers capture
sophisticated studies reflexively. At its worst, however, image-based visu e u
. and mterv'lews amongst other methods.
theIr, data with surveys' questl'onnalres
material critically and d photographs in ways which are purely illustrativ ''
1 e the former
, 1 researchers have traditiona11y captured Images
Vlsua . . But unl'k
enquiry has incorporatees even difficult to find a rationale for their inclusion. who can readIly ' appreciate the d 1' fference b etween the reartI y the� investigate
and where it is sometimnce with the history of photographic visual enqurniryal
A cursory acquaintacerns are not new as the case of the American Jou of and thelf. means of apprehending this reality, amongst the latter thIS distinction
has become confounded.
sug gests that these con otographic illustrations were a conspicuous part st
of Socio logy reve als.
Ph en 1896 and 1916 but almmo � onsider, by way of example, one of the most mfluential . works of visual
articles published in the journ al's early years b etwe
or of the journat Albiogyn all,
S � oclOlogy, Wagner's (1979) edited c 11 t' n 7ages of Information. The book
overnight they disa pp eare d. The th en edit
was necessary if sociolo towwas to ��cludes, inter alia, photographicallY�ll��:�ate research reports on the social
lfe of tramps or vagrants the wO k a tlc ' es a d cliente�e of a beauty parlour,
an inner-city ethnic com:nunity ;n �:c�me, an� pedestnans negotiating their
tion of such i mages ad
argued that the elim
mate
ina
u rishn ess , not to say quackery, and advance 9: 132r).
move 'out of a Stasz, 197
re' (Smalt 1905: 637, quoted in
responsible scientific procedu
251

250
MICHAEL EMMISON Q VISUAL DATA
THE CON CEPTUA LIZA TIO N
AND ANA LYS IS OF VIS UAL
DATA
passage on wind-swept streets. Collecting these disparate phenomena finishing touches to a model ma
n in a 'NeolithIC" Sun W. rsl .
together in this way suggests they have something in common: that in some has a white lab coat, dark suit � llp diora ma'; the artist
and tie and Sflffly pokes WIt
way each cumulatively advances our understanding of social life. But this is man' who is slightly larger '
, naked and somehow more
h his brush the 'Neolithic
not the case: the only unifying theme in each of these essays is the reliance (Born, 1998: 235) human than the 'artist' .
on photography as a means of recording and displaying information. The
apparent unity these phenomena have as a consequence of being collected
is thus entirely spurious. Whatever utility they might have as data for social Born concludes her analysis by sugges
tin that muse� m ph�tographic rec
form part of the museum',s permanen Ynal or ambIguous status and should
science does not stem from their appearance in photographs but from their need to be rescued from their cur ren t ords
lim
characteristics as objects in their own right.
Where the use of photographs, qua photographs, appears most suited to rep. resentational labour publicly aVat d'lsplay
I . ' Ma. king tl1e museum's hidden
a bi e· m thIS way
'1,
play a part in visual enquiry is in the more critical or reflexive forms of analysis WIth a move towards 'the p ost-mo d ern wou ld be consistent
museum as a mausoleum 'ded'lca te d to' . mu seu rn' an d tl1e rej.ection of the
j he ple
that have characterized the best of the representational and documentary .
obJec· ts and narratives from a t0tar ' cmg togeth ' er from fragmentary
approaches. Commentators on visual methodology invariably cite the well­
known but now relatively dated work of Bateson and Mead (1942) or Goffman Tl1e d ocumentary and rep "
Ism g h'
lsto ry , (Bo n) ' 1998 ,· 239) .
. ns
(1979) as exemplary illustrations of what can be achieved, but a number of also raised by Stuart Hall in hiresd'entcationa . n l d' Im en SlO of ph otography are
more recent studies display an equivalent sensitivity towards the use of visual attendi�g the photographic re�or� �;,,��� h� f the p robl ems of interpretation
UK durmg the 195 0s Pall's sto ry of bla ck set
materials. One of the most intriguing is Born's (1998) analysis of an exhibition , .argument IS. ' that m any 0f, the phototlement in the
of museum photography: 'Camera Obscured: Photographic Documentation could serve as documentary eVl i
graphs which
and the Public Museum'. The photographs Born analyses are part of the archival 'rnad e a public appearance i� the f'd1�enc e 0f po . st-war migra t'IOn 1lave already
Id 0f rep esentation' and
record of the in-house photographic studios of several major European and will have already acquired meanmg � as a consequence
. .
pOSItIonmg within the discourses of s and mflections through thelr. earlier
.
studio or magazine colour suppIemeth n;ws-P1l0to .agency, the photographic
North American public museums which had been collected between 1865
and 1960. The photographers' task had been to record the varied activities n� . or f1al1 thIS means that:
integral to the running of these institutions essentially the construction,
renovation, transportation of the displays. The images she examined offered It is difficult, if not by now
impossiblc, to reca,rtur ' e the car her .
a fascinating insight into the 'labour' of the museum as a working institution, these photographs. In any eve meanings of
nt� the se�rch for theIr 'essent
founding moment of meani . ial Truth' an original
mu1 hac
its social relations and hierarchies, as well as its wider social and cultural . .
ng IS an 1llUSlOn The phot
ograp hs are essentially
centual in meaning No such
significance. But the photographic images of the museum practices also engen­ untouched by the codes �nd prevI.�usly' natural moment
dered some epistemological puzzles. As Born notes, there was a 'baffling, '
transc�ndmg hIs . <
.
SOCIaI rel atlOns of pro ducr
torical time, exists . . . . Black . lO� an
of true meaning
d readm ' g, and.'
endless regress' to her analysis: �
WI'11
explOSIve little 'documents', hi sto lan s, esp eCIa lly, handling these
narrow passage which separa have to steer theIr Way through the mcr .
to interpret/ represent a representation (the exhibition) of a representation (the tes the old" Sc�II a 0f 'd ocu me easingly
' 1Ishc avant-gardism'. (I-Ia
from the new Charybdis of a ntary realism as Truth'
museum photographers' texts) of a representation (the institutionalised represen­ too-s Imp
ll, 199 1: 1 52-3)
tations of the museum). (Born, 1998: 224)
Whereas Born and Hall bot . ues
generated by largely anon mh�udea :r
l with ISS of representation in images
Born argues that in their exhibited form the museum photographs clearly or un k �ow n ph otographers, Peter Hamilton
exceed their original documentary function. Collectively, the images present (997) focuses his energi!s o 1e
of members of the Fre
� U \ m ore fam iliar photo graphic work
a 'realist meditation' on the infrastructure of the museum's own 'heroic nch humam. s sc o I, me n suc h as Robert isneau
and Henri Cartier-Bresson wh �
realism' a 'realism squared'. One way in which this was manifest lay in
World War until the Ia�e 195 ° were a�tlve from the end of theDoSec ond
the appropriation of surrealist codes - a play with scale, camera angle or 0s . F or H amI l ton the essence 0f. th'
incongruous juxtaposition in many of the images. One photograph from its attempt to capture a quintess IS wo rk wa s
;he French humanists _ picturesential 'Frenc. hness " .. � he Im . age
s produced by
classes populaire'
the 1920s depicts an 'elephant being removed for renovation'. The stuffed of street lIfe, fam
elephant, its trunk raised skyward, is shown on the back of a horse-drawn cart cO�munity, sol
served to ce1ebrate everyday lifeIlIeans, children, lovers' the
_

being slowly wheeled through the inner-city streets of London. Themes from ida . The French hu�al�. st ph d promote themes of
natural history feature in a number of other archival images: vanety of techniquesrity oto gra
to achieve these goa s. erhaps the phers employed a
the device of pOSitioni . we most important was
of the ord'ma ry
ng the VIe r 0f the scenes as a seIf'-sa me
classes wh had h d upon the activity as partme mber
A museum worker nestling among giant leaves composes a giant millipede on a
staged set while arms reach up from below and out of frame. An 'artist' puts the OWn daily routines. In 'post-o str
�
uc�ur:�� ' 1anguage the photograph their of
s created
252
253
$
THE CONCEPTUALIZATION AND ANALYSIS OF VISUAL DATA

VISUAL DA TA
MICH AEL EMM ISON

vie wers from which their pref erred ",:hic� includes photographs and also billboards, cartoons, advertisements,
the ir
particular 'subjec t-po siti ons ' fo r
milt on's ove
im ages,
rall argument is that such y photo­
dIrectIOnal signs, maps, instructional. diagram
feature of social life. But rather th an Imposmg . saand so on - are a constituent
spurio us um' ty upon them -,
.
deri ve d. Ha
readings could be idely distributed and promoted as documentar rating 1 researchers have tended t0 d 0 _ 1t . IS. necessary to look specifically at
which were initially wnes such as Life or Paris Match, served as an integ the
as Vlsua
how each of these forms operate ' e d �� life �nd t.he analytical techniques
journalism in magazi French nation which had been traumatized by
force to reunify the occupation and collaboration.
;
they demand. The addition of s �t�� :r�Sl erahons m the conceptualization
of visual data' however 0 ens �r new vlstas whi�st simultaneously allowing
experience of wartime Hall and Hamilton provides exemplary case studies more fruitful theoreti�;l c�nne t ons t0 be establIshed . ...
The work of B orn, e-based visual enquiry. Unlike 'Water and Class' one 0 f these would be to point to closer afflmhes between traditional
of the potential of ibey mag phy to
on d a dub ious illustrative use of photo gra in ethnography and 'visual enquIry ' ' . Nearly twenty y�ars ago S bmson . (1986)
each writer mo ves
th eor etic al poin ts abo ut m eaning and represele thtion
nta
rough noted that questions of place d space had been VIrtually ignored in most
qualitative research, to such a�:-xtent tl1at any. talk of partlClpant 0bservation
make important cal manifestations whichtare only availab "
their cultural and historiely the y de mo nstr ate wha can be achieved when the IS. mls. 1ead'mg: most ethnography. IS. about bstel1ll1'g . rather than looking .
visual means. C olle ctiv
d i s h arn esse d to a po werful theoretical imagimnation. of Stimson developed his call for a Visua 1 ethnography bY d escn'bmg' in detail
photographic r ecor
e fin al com men ts ab out this 'representational' ode y idio­ t11e room m ' which the General M ed'I�a1 CouncIl. (GMC) in the UK holds its
N evertheless, s om riabl
e studies are inva gical or
rder. The first is that thesum disciplinary hearings ' Most . 0 f us WI ll never ha ve occaSIOn ' to attend the
visual enquiry are in othat the re is no ob vious <c ulative methodolo regu1atory activities of profess'Ion�I �rouPs lIke . the GMC but we can readily
ext ent fs', self­
syncratic to the
to be dra wn. Such studies appear to be 'one-of ring appreciate from Stimson's desc np t IOn how the room can be vIewed, ' not
conceptual less
o
ons
rts, fasc inat ing or illu min ating in their own right conoffe
but
d is that .
SImp 1y as a place where the hearings are conducted, but a constituent element
contained rep
neo ph yte rese arc her as t o how to proceed . The seual research of the hearing itself . All aspects 0f the room - the oak. pane11'mg, leather
no guide to the petuate the narrow assimag umption that vis ' hIgh ' ceilings, the glass-fronted bookcases
they still serve to per e phot ogra phic e. In the remainder of the
chaIrs,
the tables, .the presence of a um'forme d commIssio . ' the spatIal arrangements of
nalre
' - he argues, convey
th e use of th yond the
is equivalen t t
rn
o
to my ea rlie r arg um en t about the need to move be alte rnative the formahty and solemnity of the occaSIOn: .
chapter I retu
ion al a naly sis of the imag e and to embrace not only con ce ption of
r epresentat
dim ens ion al d ata, bu t a wider, more inclusive, !his is a room in which serious matters are. discus sed.. the room has a presence that
forms of two- IS forced upon our consciousness . Th'IS IS a room that' even when unoccupied '
visual enquiry . .
Impresses on the visitor a solemn d emeanour and subdued speech . When occupied
. ' now fannal, carefullY spaken, and a matter
f

' Its
1' t retams solemnity' and speech IS
E , PLACE AND for the public record. (Stimson, 1 986: 643)
E VIS UAL: S PAC
R ETHI N KI N G TH
D I M E N S I O N ALI
TY
Stimson concludes his discussion of the GMC dIsciphnary . . . tribunal by
p Smith and I introduceout d the idea of
hin g th e Vi sual
In Res earc
(200 0) Phili hetorically posing the question of how successful he hearings would be
'dimensio nal ity' as a co re org aniz ing principle for thinking ab t isthe
men th e
differ­
point
:
�� �
If they were conducted in a radicall� dt. erent . �rclutectural space such as
of vis ual inf orm atio n. Underpinning this argu ence the various a M�Donald' s restaurant. In contrast o tradItIOn and permanence of the
ent forms al. That is, spatial conside rations influ . linary setting everythin about the fast-food restaurant signifies
dlscI
that visual is also spativisu
of al d ata tha t are ava ilab le for analysis, and spatial
meaning, or the
�
transIence and informality Fu ��s11mgs �
. an equipment are plastic, vinyl and
polystyrene; lengthy stay� ar:�lscourage through uncomfortable seating'
categories or typ es
tio ns also ent er i nto the ways we think about thevent s, which con­
considera The objects, people and e tered
the se i tem s as data. in isolation . . and proximity to the k1' tCh en areas ' 'ConversatIOn
noise ' here, for customers
relevance, of
mat eria ls for visua l analysis, are not encoun erve the myriad IS mformal. For staff it is rehearsed and repetitive. ' . .' . S. peec h here will not'
Inake l11story.
stitute the ra w we obs It would be difficult t0 conduct a dIsClplmary hearing in this
but rather in specific contexts. F or the most part patia l existence and it is thise
'
. (p. 650).
nt as also hav ing a s settmg'
features of our environme much of their socio-cultuvin ral significanc
which se rves as the me ans wh ereby
und erstood as ha g more than �timson' s analysis of the GMC room illustrates perfectly how visual enquiry
n sho rt, m ust be h1Ch is not dependent upon the P hoto graph can be conducted. By not
is imparted. Visual data , i
al c om pon ent whic h its representation in the :v
Including photographs in hIS' artIC ' Ie two important consequences have
just the two-dimensugg ion
ests. fallowed . The first is that his wOrk h as not been seen as a contribution to 'the
photog raph ic ima ge s
earc h in this mo re incl usiv e form does not, of courorse,y field' by self-proclaimed Vlsua. l researchers. But, second thelr' ab sence also
Thinking of visual res play no part. Two-dimensional images - a categ
f

mean that photographs 255

254
MICHAEL EMMISON VISUAL DA TA
THE CO NCEPT UAL
IZATION AN D ANALYSIS OF
VISUAL DATA
eliminates the epistemological confusions over the status of data in visual most appropriate analytical tools,
enquiry referred to earlier. Stimson's data are the actual objects encountered For
the value of this perspectexam PIe, S� ock and Anderson
in the room and its overall spatial configuration, It is the inferences that he
demonstr
,sIgns m , a ate " m theIrarr
Ive examm (1979
' atr, on of directional)
medical school complex, S"
' IogIc , Igns themselves, they
draws from encountering these as phenomena in their own right, not via their no SOClO ' al mte rest: rather 1't IS' the wm ays m' wh'ICh th ey ardec lare' are of
representation in an image, which are central to his methodology , In terms
of the framework being proposed here, Stimson is analysing both 'three­
and the ways these interpretations
deserve ,
0 '
attention, Sharrock and Andcca�Io� practical cour ses of
e
act
int
io
erpreted
n which
dimensional' and 'lived' visual data. I look shortly in more detail at these d'IrectlOna l sign s by shoWingehow erson IIJ ustrate thelr analytic intere .
concepts and offer further examples of research that could be conducted on they were located might conceivab ord'mary user,' Of th e premises in w.hich
' an st in
these topics , " . around, Sharrock lY use th ese sIgns m , ,
wc.y a nd Anderson �;gu� tha ' n egohat
.

mg ir
an irremediably practical a nd local , t the mterpretation of sigthe
'
but in particular contexts and in part' er, 19ns are not observed in isolans
m a
tio
is
TWO= D I MENSIONAl VISUAL DATA a b ' ld mg
HI ' IS, quite literally, a proce U ar sequences Find'mg our way aroundn
, IC 1
may need to follow the directions ofdural affair' V'lSI tors to the medical school
"

But let us relurn first to two-dimensional material for there is a need here to at thelr' deSlre ' d destination s, In th'IS pseveraI m ' formation SIg ' ns bef
, alth ough s'I gns are enorceouarriving
think of additional possibilities for investigation, Although visual researchers one at- ,a hme ,
' , the sequences in wh' roc es s,
ntered
have paid a great deal of attention to cultural and semiotic analyses of texts convey�. ng 111 ' formation about places ICorh these occur are equa11y Imp ' orta n t in
such as newspaper and magazine photographs and advertisements, there One Important lesson to be draw locales'
has been a general neglect of other equally available media material such as is the tacit 'work' tha t readers n fro m Shar�ock and Anderso
n's analysis
cartoons or comic strips, One advantage of these latter materials is that it
is generally possible to obtain lengthy historical records which can be mined
t
mus perform 111
detal'1 0f the journey that m ' be
_ or m or
' way-finding Pedestna" n e preCIs ely u sen,' f SI' gn systems
SIgns that is, can never pr
- 0
ovide every
for evidence of changing social and cultural norms, For example, elsewhere sense reasoning of the sign uust tal(en bu t 111ste
ser t0 suppIY the 'missing"
" ad rely up
on th e common-
(Emmison, 1986; Emmison and McHoul, 1987) I have demonstrated that In- terestmg ' to compare the for mformation,
. n direc Bu t it
cartoons which represent economic categories and themes can yield insights which are erected for motorI'smat ts P
of Ped�stna tion signs with thoseis
e d es tnans may mls' 111terpret
into the changing assumptions about the workings of the economy or the be farced to retrace their steps as
0
' .

matters f course, but signs and

relation between the economic and political realms which are not available to rect'f1 y a mI' ssed turn whe n 1' t '
IS a d'f
1 fer ent matter
through the conventional print record of the business pages, Prior to the 19405 P, edestrians can readily modify thonelr: ro IS' at the wheel of a car
o n a fr eew ay'
111 th �Ir
there was no idea of 'the economy' as an aggregate entity either linguistically ' tracks generally without ute s' make abrupt t ns or
, an y fea .
r that s ch vltIes will, �� f stop dead
or visually, Cartoons featuring economic messages were confined to an older collISIOns with their fellows, but
notion of 'economy' as wise or prudent expenditure. But once the modern dfl'vers, Wha t we find these are not opt!�onsach w h '
IC
Occasion
the
" n is tha t these contrast'ng 'I h a re av ail able to
f d'mg one's way a � , OgI' CS of navigation'
conception of 'the economy' appears - a process which is bound up with the
Keynesian 'revolution' in economic thinking - cartoons invariably feature
- In

by car - are reflected in differe

round on foot as fPose 0 d to f111d111g one's way
around
the economic system and its constituent components as an entity, thing or even only general autlines of wher nt sign s�:. e: Wher�as pedest ians are
a person , Inflation becomes a 'dragon' to be conquered by the would-be .
route, motonsts are invariably e to P roc d d must work out the detagiven
;
St George treasurer or chancellor. Anthropomorphic representations of sick as t IlOW to proceed, pres gI'V e n mu ch more exP l'ICI't vIs' ual info ils en
a

of recoverable errors that um abl with th ' ntlO

e 111te , n rmation
or ailing currency or the entire economy become commonplace, Importantly, pedestna.Yns rou , of aV Oid in g
key traffic d'Irectio . The re the kind
even when monetarist economic doctrine gained ascendancy and the role nal signs tend t be ' co tI" cnely engage m, sul t is that
of the government in economic affairs was downgraded, cartoonists still Up-coming road system is reprodu�ed : m 111 ,that the physical layout of an
persisted with these 'Keynesian' representational forms, turning them to s a constItuent feature of
the sign itself,
humorous or ironic advantage in their satirical endeavours ,
A great deal of 'two-dimensional' visual data we encounter in everyday TH RE E- DIM
EN SIO NAL VIS U
AL DATA
life - maps, directional signs, assembly guides, traffic regulations and the like
;h methodological investigation 0f tW
provide us with not simply information, but information which is tO,be
incorporated into practical routines, That is, this is material which, unl1ke \l�et�Itthno
draws attention to their pra
s nctIy spea
, nal
�-dl'me�slO texts, to the extent
texts such as advertisements or cartoons, is specifically designed to be used king, ctical use
foreshadowed 0 of th�s�111 partIcular environments has
by actors in the accomplishment of their goals, For these data a different set shall refer to nalytical possibilities in
Under thi� b�oad hew'
W e can loc
as 'three-dimen na�; , hat i
of theoretical coordinates, those offered by ethnomethodology, provide the ate the objects of matsio a
erial cu��� wh�c operate
' � ad ing
as sIgmfiers in social
256

257
MICHAEL EMMISON * VISUAL DATA
THE CONC EPT UALI ZATIO
N A ND ANA LYSIS OF
VI SUAL DATA
life. These range from those of everyday life encountered in the home and discuss the ite aesthetic or his torical sig "
which carry personal meanings to those in public spaces, such as statues or
monuments, which represent official public discourses. Although such forms perhaps even m's to th e point of p Ia ' g th
mfIcance, do they photograph it
em
of data can be analysed in traditional semiotic terms, they are also implicated research in Brisbane, the state ca i��� of Qu selves in the frame? Informal
the pre. dominantly masculine P olI:1t·IClan . s explo eensland, suggests that it is n ot
in human actions. Stimson was making much the same point in his discussion o ICla1 culture recognizes whoICh are 0. f mos
ff' ' rers and gen. .erals. which the
of the GMC room, although his analysis of how the activities constitutive of ut t
b h.e assorted animal figures which have t reI eva nce to VIsItors to the city,
the disciplinary hearing were shaped by its material and spatial configuration Object-centred vI'sual. enqmry . has ob�IO. US be en er ected .
did not make these methodological differences explicit. Barthes, however, does an older - and thes e days somewhat. neg ected and unfashio met h odolo gical affinities with
just this in his famous essay on the Citroen DS. We have become so accustomed Bocbl research' the use unobtrUSlve or non-reactIv. e m eana ble .'- branch of
to the cultural studies' semiotic interpretation of advertising texts that it et al., 1966) . The she,er ofvisib . sur es (e.g. Webb
is easy to forget that the objects or artefacts which figure in these texts also possible to explore social life ilit f m�ny kmds of objects m eans that it is
carry meanings. Much of Barthes' achievement was to spell out that myth­ for many kinds of object-basedcoreS�'e�atlych wecae cause �espondents are not req uired
ologies were not only located in the representational practices of advertising of normative responding prOVI�d mg : n Clrcumvent the usualprobie
campaigns but actually embodied in objects and activities. In the case of the answer. This may be particularly u�efutll1em' researcher a sOClally acceptablem
_
.
Citroen DS, Barthes likens the car to 'the great Gothic cathedrals'; he empha­
sizes the smoothness and shape of the car, and suggests its curvaceous glass
and deviance
obtrusive researorchurba n disorder" T ere IS. areswell- earching field s such as crime
contributes to a light, spiritual quality: 'Here the glass surfaces are not the form of 'traces' wand hl'ch has utI l'lzed visua1 m. foresmtablis '
atIO n
hed
111
branch. of un-
. the
windows, openings pierced in a dark shell; they are vast walls of air and space, vand. a1·Ism, and so on. PerhClCaps , cretl. ' ons ' rubb'Ish, litter, graff'Ib,. " VISIb. sele fields in
.
st well known IS. tl1e 'broken··winsigns of
c

with the curvature, the spread and the brilliance of soap bubbles' (1973: 89). theSlS CW·Ilson and Kelling, 1982th) we mo dows
But Barthes also demonstrates the value in going beyond the object itself h'IC
- such as the existence of unrepaired wm. dues that SIgnh arg · s f public inciVilities'
0
to look at the responses that people make when they encounter it in its in streets - d to attract further cnm . e becaows, abandoned cars or drinking
pristine state: th�t pol'Ice ten and residents alike do not c�e aboputtenfla1 ffenders assume
use0 0
� elgh bo urhoods . ies on these tO ICS . not needthe character of these
In the exhibition halls, the car on show is explored with an intense, amorous Impr essionistic. FStudor e xa m Ie S am �
son
0
a nd to be small scale or
studiousness: it is the great tactile phase of discovery. . . . The bodywork, the lines
un derto
. 1vedokcolla study ' P Ra ude nbus
of the so�rces and consequences 0f pubI'h (.999) recently
of union are touched, the upholstery palpated, the seats tried, the doors caressed,
mvo ectin g information f o o;:r;3,000 street egmICedIsorder which
the cushions fondled; before the wheel, one pretends to drive with one' s whole
. .
body. The object here is totally prostituted, appropriated . . (1973: 90) Using what they referr to as th: �l:ne 0 of systemat�Ic socialntsobsin Chicago.
and 'taking seriously thede ide at VISua1 cues matter' (p 605), Sampson ervation'
Generalizing, we can suggest that one of the primary advantages of objects R audenbush devised a projecta wthhich involved dn"vmg a Uhht ' . . y vehicle and
or artefacts for visual enquiry is that they offer a greater range of possibilities down. .every street in 196 Ch'lcago census
than two-dimensional data for inference making about social and cultural ac.t·IVlt!es and physical features of each sidtracts and VI'de� recording theslosocwly
behaviour and processes. There is not a great deal we can learn about 'behav­ thIS voluminous and permanent - VIs. e of the street SImultaneously. Froial
iour' from observing people reading or watching television (but see Morley, co�p1·lcated scales to capture the
_
ual record they were able to devisme
1986) but observing what people do with objects is much more promising. ScOCIal disorder. The items they coler;�men��uedOe�absence of both physical and
Often the twin strategies of 'decoding' and behavioural inference making can ars, con doms, syringes public int����
drug selling . lOn, . strgarbee �
age, graffiti, abandoned
fi ghting or arguing, and
be utilized in the same research site. For example, simply noting the placement
and the gender, racial or other demographic characteristics of the statues are not relev.anThet inthethisorcon
etical underp gs and statIstIcal
text' l't IS' mentIOn
.
. ed only t0 details of the proJ'ect
or monuments which are located in the urban environment can tell us some­ of the enormous ProVI' de some indication
potential both quan1 atIv' e and qua
thing about the values or priorities associated with men and women in the be aSSOciated with visua l enquiry.
_

3 ntitative - which can

official civic culture or national narratives (Bulbeck, 1992; Inglis, 1987). But
observing what people actually do when in the presence of such monuments
may give us a clearer idea about the attitudes or values of ordinary people LIVE D VIS
U AL DATA
(see, for example, Wagner-Pacifici and Schwartz, 1991). Indicators of the
cultural significance of particular monuments can be gauged by the simple It neverth .
task of recording the degree or intensity of contact which people make when :cord haseemp
Iess remams the case that ��:� se, �rc� .
loyed interpretive : �hlch utilizes the visual
encountering these objects. Do they stop to read information plaques, do they e turn to consider the conceptor q Ive . ec l11ques. This is evident as
of Tlved VIsual data'. From exploring
258

259
MICHAEL EMMISON " VISUAL DATA
THE CONC EPTU ALIZATIO
N AND ANALYSIS OF
VIS UAL DATA
. ... m
the posslblhtles . the use of three-dimensional data it is a short step to the
. lS. the places and settings the actual .. Visitors are sensitive to the
next 'higher' analytical le� el, w�lch peripherally, and m onitor each presence of others, both adjacently and
environments or I oca1es ·- m WhICh h umans conduct their lives. In thinking
_

use of the available space aroundother's actions so that an ordered

of this as 'lived visual data' we �ur� our attention to such matters as the
observable movements, of people . m0ftIme and space, to questions of Visibility
and invisibility and the patternmg . zon es and activities. A good dea1 0f t11e
contemporary research on these tOpICS has revisited the classic ' discussions
of the flaneur the detached or refleXlv ' �. urban spectator first introduced. by
. s of Simmel on the blase attitude
_
temporal
.. Exhibits do not app ear to hav exhibits is m aintained.
attractiveness or novelty of a e a 'uniform stopping p ower'; rather the
activities and conduct of others.given exhibit is often determined by the
Collectively, wha t this rese h is

the French poet. Baudelaire, �r the wn mg factors such as co-presence, parc

erip
pointing to is the crucial imp
ortance of
through WhlC h the metropohtan dwe ller copes with the sensory hera
conduct of visitors to such locales. l awareness and the like iri shaping the
. . 1
. overloa
. 0d
of urban lIfe. However, thIS wor ( has aI so benefited from the mclusIOn . andf and between both companions a Visual communication, in all its richness
more contemporary �nalytical framewor1(s. Some of the most interestmg the actual architectu ral layout andnd strangers, app ears to be as
significant as
innovative research 111 .these fle ld s has been carried out in locales suc h as
'

are encountered, experienced and design in shaping precisely how exhibits

museums and art ga11en�s ' Museum pl anners and designers (e.g. Dean, 1994) appreciated . '
have long been aware 0 the ability of the physical layout to influence both
the flow of 'traffic' and the natu�e of the Iearl11'n ex erience To date, however,
' ?s from thIS research hav�be�n relati�ely 'broad-brush' .
CO NC LU SIO N
the majority 0f f111dm . e�uaI peoPIe will invariably turn to the ng . ht
For example, other thmgs, bemg If we move away from the common
1
.
on entenng the museum,' people spend. more fme ' at displays at the start of sens e equation of visual research
an image-based a ctivity and as purely
an exhibition than the end; people genera 11y dislike entering areas WIt. hout embr
possible to regard many aspects of twa ce the claims being advanced here it is
visible exits. of the major figures in this discip entieth-century social science, and many
More recently a much �or� 'fme- ained' appreciation of museum visitor visual research. Visual research, inline, as contributing to the development of
behaviour has emerged m .t e wo;f .of researchers drawing upon ethno­ alized speciality that it appears to short, need no longer remain the margin­
methodology and conv�rsat�on a�a �SlS. Using video recordings of naturally Thinking of visual research as the be in the minds of most social scientists.
occurring visitor behaVIOur m ga erles and museums, vom Lehn et al. (2001; facilitate important conceptual constudy of the seen and the observable can
see also Heath et aI ., 2002) have l 00 k ed specifically at how exhibits . are as a domain of enquiry and th nections to be made between 'the visual'
encountered and exper'1enced an d th e mu t ua 1 conduct and collaboratIOn on theorists alike who might not oe w ork of many classical and contemporary
the part of both compal11. ons. and strangers w h'lCh this entails. Such researc . hr field . Several examples have alretherwise be regarded as contributing to this
draws upon the long-s�andmg s�mbolic interactionist interest in behavlO � sion. Simmel is one the m ost obviady been allud ed to in the foregoing discus­
in public places, but It add� t e c�mpl I'cating factor of how conduct IS explicitly addressed the issue ous candidates for inclusion: indeed SimmeI
inextricably embe dded in the . lmme . dlap tesed eco logy and the material realities at and the significance of mutual such as the obs ervability of social interaction
hand. The research questIOns bemg here concern precise1y how the published in 1908. But there arglances in a famous essay (Simmel, 1921) first
objects and artefacts the �xhibi 'tam � . �m and
. f:rn such installations are both given
_
and cultural enquiry Who, it can e m any other figur es in the history of social
sense by people in i�tera�h�n an w' m 'a greatobjects reflexively inform or using visual methodologie be argued, have been thinking abou t the visual
the production of
. mt mtellIglble con
. naturally occurnng � u � t . Like deal of ethnomethod­ explicitly or self-consciously ass in their work, although not necessarily as
ological enqUIry � b e h a vl' our, this research defies easy been concerned with rea ding Simmel. For examp le, to the extent that it has
0f the
. fu11 explication detailed consideration
summary and reqUIres for ItS of urban research can be conthe urban landscape then much of the tradition
actual sequences 0f behaviour. that are recorded (see also Heath' this volume). approaches to the social organstrued as a form of visual research. But other
These conSI" d �rations notwlthstan d'mg, a brief summary of some 0f the As Prior has demonstrated, ization of the built environment come to mind.
principal findmgs to emerge from this research is as follows: than a sociological geo a case can be made for 'a SOCiology of space rather
the spatial relationshipsgraphy' 0 988: 87) through fOcusing not so much on
�� �U ��U l s in pairs or family groups appear not to
. a and the like characteristicwhich exist between bUildings, settlement patterns
encounter eXh.lblts as m 1:1 U � ' . rather the experience is a consequence
It Visitors to gal�e�les
bUildings - upon their of urban landscapes, but on the internal features of
of the interactIOnal dynamIC �f �:i� a member of a pair. this field offers a faS architectural configurations. Prior's own research in
It Social interaction also dete:�mes t:e I ength of time parties spend at each Cinating account of how the
�sylums can be used to illustrate chan design of hospita ls and
exhibit as well as their deCISIOn to move on. Ing disease, ins ging institutional assumptio ns concern-
anity and the most appropriate
forms of treatment and contr ol.
260

261
MICHAEL EMMISON VISUAL DATA
THE CON CEPTUA LIZA TIO N
AND ANA LYS IS O F VIS UAL
DATA
Last, but not least, there are those contemporary schools and writers who,
in various ways, have focussed more upon the issue of the visibility of human
conduct and the opportunities this provides. Here we can mention the
tradition inspired by Foucault's example of surveillance, the panopticon and
the gaze, and how the visual can be linked to themes of power and control
(see for example Shearing and Stenning, 1995). Finally, of course, there is
the tradition of naturalist observation of social life which connects Simmel
with Goffman, his intellectual heir. Many of Goffman's most important
concepts are specifically designed to illustrate observable aspects of social con­
duct. Indeed it is something of a puzzle why Goffman relied on purely verbal
accounts of the interactions he so painstakingly documented. From Goffman
it is but a short step to the field of proxemics and the work of Hall (1966), who
did employ photographic illustrations and diagrams to visualize his work.
A longer, and possibly more contentious, route also connects Goffman to the
ethnomethodological and conversation analysis tradition where, as we have
seen, the video recording of naturally occurring interaction is emerging as an
increasingly required empirical component.
Photographic images - in either their representational or informational form
will no doubt continue to figure in visual enquiry, but only when researchers
come to appreciate the value of direct observation of the social world,
harnessed with a powerful theoretical imagination, will visual research come
to enjoy the centrality throughout the social and cultural fields which it
deserves.

N OTES

1. We cannot, of course, let the example pass without a comment on the caption.
Without this the images - regardless of the order in which they appear are almost
indecipherable. Barthes' (1977: 39) point about the need for a verbal text (a caption)
'to counter the terror of uncertain signs' has never been more effectively illustrated.
2. The list continues to expand almost monthly and the following guide to the field
will almost certainly have been superseded by the time of this chapter's publication. REF ER E N C E S
Books devoted to these themes which have recently appeared include Banks (2001),
Barnard (2001), Evans and Hall (1999), Heywood and Sandywell (1999), van Alexander V.D. (1994) 'The
lmage of Children in M agaz '
Leeuwen and Jewitt (2000), Mirzoeff (1998), Pink (200la), Prosser (1998), Sturken to 1990:, Communicatio n Resea rne AdvertIse .
ments From 1905
Banks, rch, 21 (6): 742-65 .
and Cartwright (2001) and Walker and Chaplin (1997). In addition at least two new M. (20 01) Visual Methods in
Barnard, M . (200 1) Approac · soCla
' l Research. London: Sag
journals catering specifically for this field have also made their appearance, ViRual
Communication and Journal of Visual Culture, whilst a number of other existing Barthes, R. (1973 ) Mythol ;�
hes to Understan d 'n VIsu
ogies. St Albans ' Pa .
' . aI Cu ' lture. Basingsto
e.
ke: Palgrave.
; :;
tographiC Me�sag , Image, Mu
journals now carry illustrative material on a more regular basis. Barthes, R. (1977) 'The Pho .
slc Tex t. London: Fontan
.

3. In this sense Sampson and Raudenbush's research can be seen as a continuation Bateson, G. and Mead M a
New York Academ'y of
(1942) B l me
: hic ana lysis. New York;
' se Character: a photograp
albeit one with added technical complexity - of an important tradition in visual
tography and Sociology', 5 tud .
Becker, H. (1974) 'Pho Scien ce
enquiry which has draw upon quantitative methods and large-scale samples to
make important and often unexpected empirical findings. See for example Communication, (1): . m the An thropology
les
3-26. of Visual
Richardson and Kroeber (1940); Robinson (1976); and Alexander (1994). Born, G. (998) 'Pu
blic Museums, Museum Photo . .
graphy and the LImIts of Ref
[�:��
Bu al of Material Cu ltu re, 3 leXivity',
(2) : 223-54.
, C. (1992) 'Women of . .
monuments', Hecate, Substance.' the d epIchon of
18 (2): 8-22. women in Australian

262

263
MICHAEL EMMISON VISUAL DA TA THE CONCEPTUALIZATION AND ANALYSIS OF VISUAL DATA

tions of the Self: gender and Sharrock, W.W. and Anderson D . C. (1979) 'D'lrect'IOna 1 HospI' I
Da�a ', Information Design Journa
.
Clancy, S. and Dollinger, S.J. (1993 ) 'Photographic Depic ' ta S'19ns as SOCIO logical
7/8: 477-5 08. l, 1 (2): 81-94.
age differences in social connectedness', Sex Roles, Shearmg, C. and Stenning, P. (1995) 'From the Panoph. on .
ce. Lond on: Routledge.
Dean , D. (1994) Museum Exhibition: theory and practi c to DIsney World: the
,

.. . Ioglcal
my: fetish ism and the legitimation of development of discipline' in J . M unCle .
Emmison, M. (1986) 'Visualising the Econo , .
E Mel aughlin and M Langan (eds )
economic life', Theory, Culture & Society, 3 (2): 81-97
C nmmo . ' •

. Perspectives: a reader London ' Sa e

Slm�neJ, G. (1921) [1908J 'Sociology �f the Se�ses� VIsual ntera
. J

u!, A (1987) 'Draw ing on the Economy: cartoon discourse ctIOn , 111 R. Park and
E. Burgess (eds) Introduction to the Science oJ,f SOCIO. iogy. ChIca
Emmison, M. and McHo '"
� . _ . , .

s, 1 (1): 93-11 2.
and the production of a category', Cultural Studie go: University of Chicago
ching the Visua l: images, objects, contexts and Press. pp. 356-61 .
Emmison, M . and Smith, P. (2000) Resear Small, A (1905) 'A Decad e of Socio10gy' Amer/ca
interactions in social and cultural inquiry. Lond
on: Sage. : . n Journal of Sociology, 9:. 1--10.
Evans, J. and Hall, S. (eds) (1999 ) Visua l Cultu re: The Reader. London: Sage/ Open Stasz, c. (1979) 'The E ar1y H IS
. , .
' tory of VIsua l Socio
' 10gy,, m.
}. Wagner (ed.) Images of
InformatIOn: Stzll Photography in the Social Scien es. B verly HIlls, .
SOCIO1oglca l FIeldwork', Sociological
University Press. . . . � : CA: Sage. pp. 119-36.
wood Cliffs, NJ: Prentice Hall. Stimson, G. (1986) 'Place and Spac e m
Garfinkel, H. (1967) Studies in Ethnomethodology . Engle Review, 34:
< c

nts. Lond on and Basingstoke: Macmillan. 641-56.

Sturken, M. and Cartwright, L . (200·.1 ) IJraetIces
Goffman, E. (1979) Gende r Adver tiseme
Doubleday. ' ofLookmg.' an mt
Hall, E.T. (1966) The Hidden Dimension. New York: . . TOductIOn
. •

' to visual cultur e.

of post-w ar black settlement', in J. Spence Oxford: Oxford University Press.
Hall, S. (1991) 'Reconstruction Work: images van Leeuwen, T. and }ewitt, C. ( 2000) Handbook 0f Visua .
Dome stic Photog raphy, London: Virago. "

LIm
of
' dmars h } (2001) 'E xh'b'
and P. Holland (eds) The Mean ings l AnalYSIS. Lond on: Sage.
e and Frenchness in post-war vom Lehn" D. Heath , C . and 1. .
Hamilton, P. (1997 ) 'Representing the Social: Franc and collaboration in museums and gall�ries', Symbolic Intera
I Itmg Interaction: cond uct
Cultu ral Representation and
humanist photography', in S. Hall (ed.) Representation: Wagner, }. (ed.) (1979) Images 0f InfiormatIOn: . ctIOn, 24 (2): 189--2 16.
Signifying Practice. London: Sage. Still Photography in the Social Sciences.
and Cleverly, J. (2002) 'Crafting Beverly Hills, CA: Sage.
Heath, c., Luff, P., vom Lehn, D., Hindmarsh, J. , Visual Communication, Wagner�Pacifici, R. and Schwartz, B. (1991) 'The V'letnam
Participation: designing ecologies, configuring experience' Veterans, Memorial',
Amerzcan Journal of Sociology, 97: 376-420.
Visua l Culture: explorations in the
1 (1): 9-33 Walker, J. and Chaplin, S. (1997) Visual Culture-. a n mtrod .
Heywood, 1. and Sandywell, B. (1 999) Interpreting uctlOn. Manchester: Manchester
.

UniverSity Press.
Anzac Australia', Daedalus, 116 (3):
hermeneutics of the visual. London: Routledge
.
War memo rials: Webb, E.}., Campbell, D.T., Schwartz, RD. and Sechrest, L. (1966) .
Inglis, K. (1987) 'Men, Wom en and Unobtruszve Measures:
non-reactzve research in the social sCienc
;
' es. CI1lcago'
�
35-58. W'II son, J . and Kelling, G. (1982) 'Broken Wi d ' Rand McN .
a11y.
London: Sage.
Mason, J. (2002) Qualitative Researching, 2nd Edition.
� �:;
" he lantzc Monthly, 249 (3): 29-38.
l Cultur e Reade r. Lond on: Routledge. Ziller, RC. (1990) Photographing the self: e h ; or 0 servm
Mirzoeff, N. (ed.) (1998) The Visua g personal orientations .
cultur al power and domes tic leisure. London: Comedia. Newbury Park, CA: Sage.
Morley, D. (1986) Famil y Television:
ecture of everyday life, 3rd Edition.
Newman, D. (2000) Sociology: exploring the archit
Thousand Oaks, CA: Pine Forge Press .
and representation in research.
Pink, S. (2001a ) Doing Visual Ethnography: images, media
gies: on video, reflexivity and
London: Sage.
Pink, S. (2001b) 'More Visualising, More Methodolo
study of spatial organization and
qualitative research', Sociologica l Revie w, 49: 586-9 9.
Prior, L. (1988 ) 'The Architecture of the Hospital: a
(1): 86-1 13.
medical knowledge', British Journal of Sociology, 39
book for qualitative researchers. London:
Prosser, J. (1998 ) Image-Based Research: a source
RoutledgeFalmer.
ries of Women's Dress Fashions:
Richardson, J. and Kroeber, AL. (1940) 'Three Centu
of the Beard: the men of the
1 11-53.
a quantitative analysis', Anthropological Records, 5 (2):
Trimm ing
Robinson, D. (1976) 'Fashions in Shaving and
al of Sociology, 81 (5): 1 133-41 .
Illustrated London News, 1842-1972', American Journ
Social Observation of Public
Sampson, R and Raudenbush, S. (1999 ) 'SystematiC
borho ods', American Journal of
Space s: a new look at disorder in urban neigh
Conversation', in P. Drew and
Sociology, 105 (3): 603-5 1.
Schegloff, E.A (1988 ) 'Goffman and the Analysis of
order. Oxford: Polity Press.
T. Wootton (eds) Erving Goffman: exploring the interaction
graph y in qualitative research',
Schwartz, D. (1989) 'Visual Ethnography: using photo
Quali tative Sociology, 12 (2): 119-54.

264 265
 ANAL YSING FACE-TO-FACE INTERACTION

is audio-visual recordings of 'naturally occurring' activities and events. We

then take an example drawn from a medical consultation and discuss the ways
in which we can begin to examine aspects of the participants' conduct.
Before considering some basic methodological issues, it is worth briefly
Analysi n g m entioning that whilst sociology has been relatively slow to exploit the
opportunities afforded by video, there is a long-standing, though largely dis­
regarded, tradition in social anthropology of 'interaction analysis' using
d recorded data of naturalistic activities. It emerged in the early 1950s in Palo
Vi deo, th e vis ual an Alto through the pioneering work of Bateson and Mead and others including
McQuown, Hockett and Ruesch, and led to extraordinary studies undertaken
Christia n He ath by Birdwhistell (1970), Scheflen (1964) and others. Coupled with the wide­
ranging essays by Goffman (1959, 1963, 1967,) the tradition also influenced
Kendon's (1991) impressive work on gesture and 'bodily comportment in
interaction and the more ethological studies by Cosnier (1978) and others.
These studies have some parallels with a very different tradition which
has led to the emergence of a distinctive body of sociological research con­
cerned with the ways in which talk and bodily conduct feature in social
. de elo ments in sociology over the past couple interaction. In particular, ethnomethodology and conversation analysis have
. ressIve
Imp provided the methodological resources which have informed the development
One of the m ost
? the burg eon
�
:� � od of empirical research con cerned with
of a growing body of sociological research which uses video recordings to
o gY �nd conversation analy org
of dec ades has ee n sis hav� m�de
talk in inte ract ion. Eth no meth o ial anlZatlOn examine the in situ organization of social actions and activities in face-to-face
·b u fI on to our un d ers.t an
d·In g of the soc interaction. For example, in a pioneering series of essays, C. Goodwin (1979,
a pro fou nd co ntn · ch they featur e in the aCCOffi-
an d t a n( and the wa. ys 111 whl s 1980, 1981) examined the ways in which the production of turn at talk, a
of lang uage
�
use . a1 acf10ns and activi·t·les . The visual and ma ter ial asp . ectI
plis hm en of so c � . it is inc rea sIn ? Y speaker's utterance, is coordinated with the gaze of the recipient, and went
InteractIOn have recelVed. les s attention and yet prachcal on to identify various devices employed by speakers to establish mutual
of social
og . d that b0dI·lY conduct' objects d artefacts are criticalg toaVthe
nlze
an . ab
all l·1·lty 0f orientation. Related research examined the organization of gesture, and in
rec . and actIVI · ·ty. The increasin
accompl1shment 0f SOCI·al act Ievant
. IOn particular the ways in which various forms of bodily conduct are used by
. ment, coup1ed with the em
Ulp erg en ce of re
le vid eo eq . listic speakers to shape the co-participation of the person(s) with whom they are
d�ng to a ?r
che ap an d rel iab . interest in the nat ura
methodological resource.s,nisalea nd t � ways In : �� �� it is acc om pli she d thr oug h speaking (see for example C. Goodwin, 1980, 1 981; M.H. Goodwin, 1980;
analysis of social in tera ctIO d. Heath, 1986). Since these beginnings we have witnessed the emergence of a
talk, bodily conduct �nd ththe e matenal. resources at h anuse video recordings of wide-ranging body of research which has used video recordings of 'naturally
nsId ers v: ays 111 w hic h we can ants occurring events' to examine visual, vocal and material aspects of a range
This chapter co .
n. to exam111e the ways in whi ch particip
con duct an d ·
111 t� rac tIO . ular It
rtic of activities in both conversational and institutional environments (see for
accomplish particul.ar actl�ltleS
11 boration with oth
pa
.

e very day . . ers . In

suggests that analytic deve �pm: t .i hin
� �: � the social sciences, namewh
ource s thr ou gh
ly, ethno­
ich we
example Goodwin 1987, 1994, 1996; Goodwin and Goodwin, 1 996a; LeBaron
and Streeck, 2000; Streeck, 1988, 1996; Hindmarsh and Pilnick, 2002; Whalen,
methodology and con versatIOn analysls' prov. ide res ual, 1995; vom Lehn et al., 2001; Hindmarsh and Heath, 2000; Heath, 2002; Heath
0 f SOCIa 1 act·on
. I and add ress the vis
e d pro d u ctio n and Luff, 2000).
can examine the s Ituat
n feat u es 0f an activity' s acco
mplishm ent. The ?cuS f
as s p k � s in which we can begmlvtoes
material as well � �
ract � on, a nd the wa
of the chapter is socI al
and
Inte
proce ure � th r h vI hich participants themsemake
enviro�m;� s pr���ce their ow
analyse the pra ctic es n actions and ANALYTIC CONSIDERATIONS

in ordinary, everyday othe rs. n t 11 S way, the ch. apter gests that sociotment
sug logy
duct of
sense of the con . . . appro ch to the ways In w h·ICh bodily compor Conversation analysis and ethnomethodology provide the resources through
can develop a dlst111ctIve ture �111 every d ay acfIVities and contribute to our which it has been possible to exploit video for sociological purposes. Before
an d Physical artefacts fea . presenting an example, it might be helpful to provide a brief overview of the
I f nti. ve Issues.
0
. �. � ketch of one or two m ore gen era1t
of a rang e f ana y lC a nd sub sta �nalytic issues which informed the research. In general, analysis has been
understanding by ro� ld tha Informed by three principal issues. Firstly, it has been concerned with the
s � �
The chapter
methodologi cal
beg
cons
i r:
Ide ratI
?
Ons w IC In � orm the analysis of video data,
267
CHRISTIAN HEATH VISUAL DATA
ANALYS ING FACE·TO-
FACE INT ERA CTIO N
detailed analysis of the local, situated accomplishment of social actions of everyday life, whilst we concerte
and activities, Secondly, it has been directed towards the investigation of the we confront. dIy accomplIsh ' the very scenes and eve
methodological resources, the practices and reasoning, through which partici­ nts
It can be argued that conversa
ments and in particular treats 'conve anal tion
pants themselves produce social actions and make sense of the actions and , On' YSlS' shares these analytic commit-
activities of others, Thirdly, conversation analysis has drawn on the emergent as a reaIm 0f sociological enquiry rsatI and its meth0d0Ioglca ' 1 lou
L�ct�res on Conversatio ndations
1:
n embodi es
and sequential character of talk in interaction to examine the ways in which Wh' Il S t h '
IS
of Sacks (992) and his colIeagues d �na1rIC InsIghts, it is p erhaps the work
a diverse range of observationS an ' .
participants orient to and make sense of each other's conduct and
locally relevant action and activities, organization of talk which has, S� eg off and Jefferson on the sequential
Ethnomethodology emerged through the pioneering studies of Harold frUlt' fu 1 contribution s to ethnomP roVlded one of the :nost WId
studies, reflected p erhaps most e eIth' �d0I, ogi' cal s�udles, It is certainly these
' e-ranging and
Garfinkel and rapidly led to the development of conversation analysis through
the innovative research of Sacks and his colleagues, Schegloff and Jefferson, of turn taking in conversati on (SXP ICl ly ln a claSSIC paper on the organiza' tion
Unlike other forms of social science enquiry, ethnomethodology and conver­ ' fl, uence on the analYSis of talk in
m ac1(s et aI 1974) wh'ICh has h
ad a profound
plmes, It should be said however, both SOCIOlogy and a range of other disci-
f

sation analysis do not provide a 'method', in the sense of a clear-cut set of " "

procedures that if followed will generate scientifically valid results or findings, WIt,, h ,Ianguage per se, bu ' that conVersatIOn anaIYSIS" IS
However, they do involve a number of critical analytic commitments which t rather d env
pnnCIpal means through which we ' es from the r ecogmt' 1On . not concerned
have provided a foundation to a substantial body of empirical studies, ac t'IVl't'les, 5chegloff and Sacks (97. p roduce. and recogn'lze SOCIa , Ithaacti
t talk is a
ons and
In Studies in Ethnomethodology Garfinkel (1967) develops a radical approach interest in talk derived from the rec 4) a,r?ued for exam ple that their original
to the analysis of human practical activity, He argues that we place the situated deve1OpI' ng a 'naturalistic diSCipli ogmhon that it prov'1ded th. e possibility of
production of social actions and activities at the forefront of the analytic action(�) �igorously, empirically ne wh'�h uld , . de�l with t�e d etails of social
agenda and treat mundane events, even physical and biological phenomena, transcnptIons of naturally OCcur ' and f� m��lY , Usmg audIO recordings and
a substantial corp us of empiric ng taII,<' conv�rsation analysis has developed
as the 'artful accomplishment' of the participants in the settings in which they
arise, Garfinkel suggests that we 'bracket' events and ask of any phenomenon reasom,ng, the comp etencies whi :\ studles whIch delineate the practIc'
ch In ' form tI1e accomplishm es and
what the methodological resources are which inform the production and range 0f conversation activitie ' ent 0 f a div erse
intelligibility of the event or activity in question, At one point, he contrasts versat'IOn I1as been increaSingly s ' In r ecen t years' the , 0, ngma1 focus on con-
talk and a rich body of work re p I ace d b
"

y a groWIng rest in institu

his analytic recommendations to those recommended by Durkheim:
orgamz, atIo , n has emerged concernedmte WI' th th e mter
tional
of events such eas, news mte , ' actio nal
1 lcaI sp eech es
po 1't' ' rvIewsf m ed'lcaI conS
Thereby, in contrast to certain versions of Durkheim that teach that the objective (see for exampI e B0den and , Ultations and
reality of social facts is sociology's fundamental prinCiple, the lesson is taken instead and Maynard, 2004), ZImmerman, 1991; Heritage
and used as a study policy, that the objective reality of social facts as on ongoing Fo� Ga:finkel the 'indexicality
Indexicahty p oints to the uni of ractical acti�r:s' IS' a fundamental concern,
attention towards the ways uene�s of an� �ChVlty or event, and draws our
accomplishment of the concerted activities of daily life, with the ordinary artful
ways of that accomplishment being by members known, used, and taken for
routin e and mundane characte; h'IChtartJcI� ants accomplish the rational
activities which informs ethn �;:r�c tcal actIOn. The uniqueness of practical
granted, is for members doing sociology, a fundamental phenomena, (Garfinkel,
1967: vii)
radical than the notion of coomet � ,og and conversation analysis is more
So, for example, in their different studies of 'suicide problem', both Garfinkel �11 too often 'context' is trea ntext t� �: r�;nd elsewhe�e in social sciences,
(1967) and Atkinson (1978) do not take the official category 'suicide' for InVoked to explain the sp ted as t:e a of local vanables w hich can be
granted and examine the variables which explain patterns in the rates of Par,t'ICUI ar occasion or circum ecTI IC C hara cter of practicaI ac "
suicide amongst particular classes of individuals, Rather, they examine the SOCIOlogical enquiry suc s' tance ' I�1deed, ven moretIVralty within some
, mter � dica forms of
practices and practical reasoning on which those responsible for the investi­ ments 111 cognitive s cience
, ' h as symb0lIC
C , act10111S' � and certainl dev
gation of equivocal deaths rely, in warrantably ascribing the category 'suicide' human conduct that such as distributed cog111t1On, elop-
whil " reta in a mo del
in a particular case, In this way, analytic attention is directed towards the st
action, presupposes' th emphaSlzmg the temporal orgamza " t1On of practicof
in situ accomplishment of particular events and activities, and in particular re�erence and the lik at shar ed mea�111gs ,
' or defI�I" honsf comm al
WIthin social life, In contra stabl If ��l�f�r bnef moments of interactio f
e r em ain on fra mes o
the resources on which individuals rely in the production of social actions
and activities, Moreover, it is argued that we take for granted or gloss the social actions and activi st' in ethn�m e o O ogy and conversation analy n
systematic ways in which we accomplish social actions and activities, th� of" the 'cor:text at ties are inseparable from or
" bett ·er, p
sis'
the objectIv
, hand ', The intel ility of a tcene, the cha art a nd p arcel
'objective order to social facts', so as to encounter the 'normal appearances e order of soci al factsligib
"f IS ongom , gY racter of the event
C
accomplished in and througl�
268

269
-
ANAL YSING FACE-TO-FACE INTERACTION

CHRI STIAN HEATH VISUAL DATA

re is 'no Unfortunately, however, there has ·

been a w Idespread tendency amongst
an d con cer ted act ion s of the participants themselves; the er research. on non-verbal comm umca. t .
IOn to assume th a t the meamng . of a
the practical
m the mo me nt- by -m om en t production of the 'objectivise aord cen ­ gesture IS mextricably tied to its phYSiCaI f?rm rather than the context in which
. .
time out' fro ive cha rac ter of practical action therefore ndations it. arises, .an assumption which IS. not unl1ke the idea that an utterance gains
of social fac ts' . Th e ref lex thodological fou
alytic attention to the meed 1 tS m eamng solely by virtue of its lexicon �r �ynta x, rather than sequential
tral concern and directs anact ivi tie s an d the achiev character of ordinary locatIon. . We might also suggest that the phYSlca enVIronment . . which actions
m
of practical act ion s an d
turn- and activities take place does not have a stable and overarching influence
events.
rsa tio n an aly sis , wi th its pri nci pal focus on interaction, the res ou rce on interaction, but rather its reI evance and sense is accomp1IS· h ed by partici-
For conve has provided an important
al org an iza tio n of tal k cti c pants within the interaction as it emerges moment by moment, In consequence,
by-turn, seq ue nti
of con tex t an d the 'indexical properties ofe action alis
pra
we need to consider the ways in whIC · h Vlsual. . conduct features '
for the an aly tic de pic tio n
example suggests that 'coer'mm unicativ
with and within talk, and to draw on the sequential orgamzatJon. .in interaction
of conduct
action' . Heritage (19 84) for con tex t ren ew ing '; 'a spe ak s con tribution is both
as a resource for the analysis of in Sl.tU socral . actIOns . and activities,
both contex t sha pe d an d particular
wi th reg ard to the loc al con figuration of activity and inute s to the Visual and tactile elem�nt s 0f h uman conduct" features 0f the local
designed act ion s, and itself inevitably contrib 84 .
enVIronment
the imme dia tel y pre ced
wh
ing
ich the ne xt act ion will be understood' by: 242).
(19 . . and
orgamzatIOn,
the like' do. not ref1. ect. the turn-by··t
characteristic of talk m mteractIOn. . Whilst urn, speak er-by·-speaker
1·t IS· spea kers who
framework in ter ms of ere each
y-s tep , seq ue nti al org ani zat ion of talk in interaction, whrec ognized ordman . . 1y gesticulate, even a single turn at taU may mvolve . a complex array
This step-b rstanding of prior and is ise . <:

ys an un de of actions produced by va nous partlClpants

pla . some perceptual range of and
, both spea k er listeners
.

subsequen t tur n bo th dis indi­

.

ard to the im me dia tel y pre ced ing action(s) (unless otherwstipula te and. others, which may stand 'WI. thm the event' (cf
with reg
vid es an im po rta nt ana lyt ic res ource. Rather than simplytheir own Goffman, 1981) · For exam I e �S suggeste� earlier, a single utterance may b�
cated ), pro particular utterances in the light of
coordinated in the cours, ! 0f 1 tS prod uctIOn with the VIsua
the meaning or significancechof ers can inspect su bse qu en t act ions in order to
person(s) to whom it is addressed' and th e spe��er may employ various
· 1 actions of the
uit ion , res ear ying
persona l
ne
int
ho w the pa rtic ipa nts the ms elves are responding to and displa gestures and the like to shape the forms 0f co-parhclpation he or she requires
at d1. fferent junctures within the. t��n' s deveI opment. The utterance, and the
determi rs' conduct.
their understanding of eacheothe ction whereby sequential orgaani zation
The do ub le- ed ged fea tur of int era me thod­ way in which it is understood IS e �ut�ome of a complex interaction that
an int egr al fea tur e of the soc ial organization of talk and l feature of includes both visual and vocal contn. u.tIOns by the participants during the
is both alysis remains a centralest and powerfu , the Vlsua
ological resource for its an , for example: very course of its production. When conSI dermg · 1, vocal and material
features of activities in interactI· on therefore it is important to conSIder the
tic res ear ch . As Sa cks et al. su gg .
conversa tio n an aly . . . '
sation of conversation that
sequentIal organization of the partIClpants' conduct even th ough next actions
ce of the turn taking organi may occur prior to next turn A . th: ��rr;Kle abov�, it is found that the
[It] is a systematic consequen
it obliges its par tici pan ts to dis pla y to eac
ir understand­
h other, in a tum's talk, the as directed
will be heard reorie�tation by a po tential r�ci i��t it ; e emergmg utterance or turn
ing of the other tum's talk
. More generally, a tum's talk
d to loc ate some other at talk IS engendered through th :peal(.er' s gestures and bodily conduct. Or,
for example, a co-participant wi� xamme so�e feature of the local environ-
iqu es are use
, un less special techn
to a pri or tur n's talk ings of other turns' talk are
. . . . But while understand
talk to which it is directed available as we ll to profes
sional analysts, ment, such as picture on a wall' a rme 0f text m a document or a dJagram
· on
tici pan ts, the y are
ure) for the analysis . . ' a participant
criterion (and a search proced derstandings of
displayed to co- par a computer screen, by virtue of the actions of a co-partIcIpant
who, through bodily comportme t' �::;� ��. lr renders the o�ject relev ant.
pro of
:
reb y pro vid ed a
�:
who are the parties' un
upied with. Since it is the
So whil e visual conduct with an
of what a tum's talk is occ nt to the ir con str uct ion of next turns, it is their IS_ not necessanly organized
tha t is rel eva se understandings .
prior turns' talk d for analysis. The display of tho on a turn-by-turn basis, we can conSI· der the ways in wh·ICh partIcipants
wa nte
understand ing s tha t are analysis of prior turns, . . .
uen t tur ns affords a resource for the intrinsic respond to each others' conduct as a wa. y of Investrgatmg how their activities
are organized with regard to the ac tIOns 0f the other . The sequentIal
seq
in the talk in sub
for pro fes sio nal ana lys es of prior turns, resources . and
and a proof procedure .
to the data themselves.
(1974: 728-9) interactlOnal organization of the con duct remains a cn·fIca I resource for
to the anaIysis of how participants thems eIves onent . to each other's action, make
d seq ue nti al org an iza tio n of interaction is also relevandut ct sense of each other's contribufIOns, and produce their own conduct.
The emergent an contextual or in situ signifGe icance of visual con
how we might consider thes of hu ma n en vir on ments. stures and othart er forms
and the physical pro pe rtie
not infrequently use ef acts
in int era cti on , pe op le l
of bodily condu ct ari
eac h
se
oth er, and it is no t us ua l for aspects of the phvitica
ys
ies.
when talking to ant within the course of soc
ial acti
environment to become relev
271

270
CHRISTIAN HEATH " VISUAL DATA
ANA LYSIN G FACE -TO-F
ACE INte RACTIO N
OBSERV I N G CASES of human conduct and loc
f others in the late nineteenthomo tion originally initiated by M
The situated character of pr � tical :,c ,lOn, nd the interest in the method- cen eybridge an d
and theoretical resources whi ch tury , Similarly, in sociology the conceptual
ological resources used by t � par l�lP �nt: themselves ' inevitably drives ethnographic work since the 1950s,have informed a substantial corpus of rich
analytic attentlOn' towards the mvesttgatlOn 0f activities and events within lend them selves to an analysis of for example Hughes (1958), do not readily
the contexts m ' W h'Ich they occur , , . . 0, , ecou
tal'1ed and repeated inspection 0f the captured on video. In contrast, howthe details of social actions and activities
accomplishment of actual aChVlt1eS pled with the analytic orientat'lOns analysis, with their commitment toever, ethnomethodology and conversation
briefly discussed ab ove, provide th e resourc es through. which researchers
, 1ar conduct, and their interest in taki the local in situ organization of human
can b egm' to identify the prachces , and reasomn ' g. through which parhcu in their Own right p rovide an anang talk and interaction seriously, as topics
events are produced and. rendered intelligible, of the opportunities afforded throlytic orientation which can take advantage
As some ethnomethodo1 o�lca ' 1
become increasingly concerne Wl't, :�aIk and interaction, it has been found
d conversation analysis researchers As stud ies of talk and interactugh video,
with more specializ ed forms of socia ion have become increa Singly
concerned
that audIO' and. aud'10-visual record mgs . prov ide useful resources, with which organiz ational or institutional do l activity, often ariSing within particular
to subject in situ practical act�, ons ��d act1" �1ties to detailed analysIs, Ongma11Y
"
necessary to augment recorded mains, it has be�n recognized that it is
the use of recorded data pnman y c0nslste d principally of telephone con­ own studies of general practicematerials with field work. So for example Our
versations ,(se� for exa�llple
. Sacks 1992 or Sacks et al " 1974) ,' however, with observation before any recordinginvolved a long p eriod of non-participant
the increasmg mterest m the VISU ' al mat�rial as well as vocal aspects of human sense of the organiz ation of certatook p lace in order to begin to assemble a
activity the use of video has �ecome mcr , eas' ingly common (see for example
treatment and using m edical records in specializ ed tasks such as diagnosis,
Goodwin, 1994, 1996 G odwm and G00dwm, , 1996b, LeBaron and Streeck,
' the emergence of more wid e-r (Heath, 1984; Heath and Luf(
2000). With
; . � Wha1en, 1995,' Hindmarsh and Heath, 2000, 2003;
2000' Streeck, 1988, 1996, especially tho se concerned with angi ng stud ies of the workplace inter
action,
Heath and Luff, 2000) ' It has been recogmze ' d that recordings of human technological environm ents such the use of tool s and artefacts in complex
activities and m "
' teraction despIte ' , thelf' l'Iml'tations, provide researchers Wlth ,
we have witnessed an increaSin as control rooms and emergency centres,
unpar�lleIed access to sOClal , :� ��n a11 wing the aspects of the comp 1eXl'tY
! g commitment to undertaking
of partIcular events to be sub) �o d�tailed and repeated scrutiny. Unlike fieldwork alongside more focu
Good win, 1996b; Hindmarsh and sed interaction analyses (Goodwin and
wide-ranging
1 , . and. quantlta
other forms 0f qua native ' t'lVe 'data' , recordings of naturall
. 'flyc Luff et a1., 2000; H eath and Luff, Pilnick 2002; Suchman, 1993; Whalen, 1995;
, not on 1Y rovide colleagues and the 'sClenil
occurring human actlvltles and even interviews d erives not 2000), The necessity to undertake observation
' , at 1 arge with access to th e raw materials on which the investi-
P
commumty forms of activities under scrutinsimp ly from the complexity of the sp
gations are b ased, but also provl'de a corpus of data which can serve a range distributed activities which feat y, bu t as a consequence of the rangeecialized
" mterests, A s Hen'tage and Atkinson suggest: of often
of theoretical and analytIc ment of the work and tasks in ure, if only momentarily, in the accomplish­
delay gathering recorded ma question, It is not unusual in such studies to
In sum' the use of recorded data serves aS' a cOntroi n the limitations and fallibilities standing of the activities in questerials Until researchers have a
passing under­
of intuition and reco11ecflOn,, it exposes the ob server0 to a wide range of interactional which feature in the accomplish tion and the various tools and technologies
'
materials and ClYcums tances and also provl' des so me guarantee that ana1Yt'Ie such settings. m ent of even the more munda
ne activities in
,
�
considerations will not anse as artef cts of intuitive idiosyncrasy, selective attent'IOn
or recollection, or experimental deslgn, (1984,, 4)
, , . , t d nstraints of video, it is surprising that THE BODY IN ACT
ION
Notwithstanding the hn:ltat1 ��� �
there has been such relahvely 1 e Int��est within sociology, and in particul�r As a way of illu ting
' eX 1 O . ting the possibilities It organization of anstra how We might begin to unpack
provides, In anthropology, film and more recenKy �ideo, has received more
P erhaps field studies and ethnography, m acti vity, aspects of the
fragment. It is drawn from a it is p erhaps useful to consider the fOllowing
attention, and yet even there, i� has been increasingly used as a med'mm �,f of the consultation as the medical consultation and arises towards the end
representa t'�on and documenta t'IOn rather than a resource for the analYSIS to write, the p atient whdoc tor begins to p repare a p rescripti
on, As he ins
" "
of social actlOns and achvlhes, Marks (1995) suggests that ethnograp,h'1C film tion, begins to teU a story still standing following the physical exabeg
o is
mina­
has been influenced by successlVe, shifts in anthropological theory smce the she ha s walking, (Details , a story that gives an example of the difficulti
beginning 0f the century, and that the absence of a suitable
' methodological and Appendix to this volum e,) of the transcrip tion systems can be found in the es
conceptual framework undermme ' d the early attempts to examine the details
272

273
 ANALYSING FACE· TO-FACE INTERACTION

VISUAL DA TA
CH RIS TIA N HE AT H
Fragment 1 . Transcript 2 and F igure 1 4. 1

Frag me nt 1 walks
up down up down up down up down
ite a prescrip tio n}
Dr: {B eg ins to wr
(eh) I we n p.
(1 .4) I an I felt so aw : :fu l � �
I was coming up the steps li: e thi all the �ay u � I felt,
I we nt go wn int o De be nh am s I fel t, (0. 4)
.
Wh en y up
( ) I was co mi ng up
P: all the wa
t h e steps I i : ke thi s
1\

ter rib ly Dr: writes

(0 .3)
P : terrib py (.) rea lly you kn ow
prescription

Dr : \J. eh yes
Th e patIent. assembles an activity, a story
.
kn ee itself ( .) yo u'v e
(0 2) the re.
go so me rhe um ati sm
.
which illustrates the problems sh e has suffered .
The!at'lent foregrounds aspects of her bodil
Dr : No ::: ( ) it' s the

of the complaint with wh ich she has

pr ov ide s an ex am ple lked up l uct, and attempts to have the body. itseit
The patient's sto ry
th e ev ents that occurred as cularwadescribes
sh e ��a�ure, then and there ' as an mterachonally'
contacted th e do cto r. It rec ou nt s
re, and in parti ."
the steps at Debenham s, a local department stono r responds relevant and ' re gar d e d feature of the �artlcl­
g sh e inc ur red . It is int ere sting to te how the dopecto rie nced with
.
pants' activities. Talk. and bod'1Y actIvIty are .
the sufferin th e su ffering the patient ex wever, rather mutuall y and reflexIvely constituted to gIve
1
.
. He ac kn ow led ge s ss' . Ho
to the story ap with the word 'terriblene ies, a sense o� the event and her su££ering .
'yeh' and the 'yes' in overerldiscussion concerning the patient's diffictult so me What IS curious pe. rhaps IS' that the patient
tells a st : : qUlreS
than encour ag e fu rth yo u' ve go
eq ue nt tu rn , 'N o:: : (.) it's the knee itself (.) and provides a the potential recipient
both to � re� �� to. look when he shows little
the doctor's su bs difficulty
rheumatism there', sim ply pinpoints the areadiof sco urages further discussion
In th is wa y the docto r interest in ternporanly abandoning the activity
diagnostic as se ssm en t.
urred by the patient eanHe d (re)aligns the in wh�ICh he. IS. engage d . 1"he d octor does not
es an d su ffe rin g inc , 2002) .
of the diffi cu lti
tow ar ds the ma na ge me nt of the problem (se ngath up the stairs look u� dunng the first part of the story where
consultation scr ibe s th e dif fic ulties she had walki ng the problems the patIent sets the scene and ind eed even as
As the pa tie nt de illustrati
at Debenhams, she wapl lks up and down on theespdoot,ctor's desk and, balancing she begins to walk up and down leaning . theon
th
e aces here hand onich she distorted her hip and leg his desk, he con t'mues doggedl to wnte
she experienced. Shstr s the way in wh prescription. Unless the patient Zan encourage
her weight, demon ate th e sta irs . Th e mo ve me nts give sense to the talkd the doctor not only to listen' but also to look
movement to actuallyey cli mb perienced an
pa ny . Th lu cid ly re ve al the problems she ex e sto ry points to and tch he� performance, then the gist of
they acco m rin g sh e in cu rred. Th the s��. Y and ItS potential
provide a vivid pictuov re of th e su ffe
me wo rk in wh ich the movements embody, cance IS l ost. In the w. a� 111 . sequential signifi-
wh'Ich the story is
es a fra
id
the difficulties and pr difficulties and suffering.
literally, the patient's e story, and for the story to achieve its local seque ntial :
told the t' ' S:Ch:Ity r�quires the doctor
to tr�nsf tr� ;� t ay 111 WhICh he participates
1

To make sense of th nce, therefore the doctor needs both to hear as well in the . teractlOn, '
and interactional significa
111

nt says and does. nduct of are as� e,ct� of the earlier part of the
as to see what the patieinc lu de s so m e de tai ls co ncerning the visual cons d patient's story which may well re��:�� , et� sensltIvity to the activity in which
A transcript which be helpful. The participants' conduct is tra cribe the doctor is engaged and her det
articipant. For example, the rest:;��� ;� aIr I we n (.) I was' may itself be
t above and be low n t ?� sta blish a more involved co­
the participants mighasp the visual conduct indicated
across th e pa ge wi th ect
yin
s
g
of
im ag es are tak en from the or igi nal. The images �n atte�pt to encou rage the practitioner �0 t��n ;owards the p atient (d,
the talk. The accompan e identity of the participants. GO dWl , 1981 ) and the sentence 'I an I felt -
a �
gr ad ed to co nc ea l th 11 , so aw .. ul coupled with the restart
n emphaSIS on 'so awful' may we11 be an attempt t0 de1ay t11e gISt
have be en de ' of the
story wh'l I e encouraging the doeto t e porar '
Ily to ab a nd on the e ion,
If so they fail, and deep into the tu�n :e atIe
' pr script
, t constructs the story so that
? �
it demands a visually attentive as weII as lIstenl11g recipient. Despite the risk,

275

274
 I

ANAL YSING FACE-TO-FACE INTERACTION

VISUAL DA TA
CHR ISTIAN H EATH "

her ,?ut
experien ce and sufferi �g tlOn
abandon his current activity to witness the difficulties that she experienced
. t not only '
I llu strates walking up the steps at Debenhams.
however, the patien ?es the doctor to ab an don writing the prescnptory The story, its impact and the doctor's ability to respond are the outcome of
successfully enc?ur� Icant p rt f the performance . In this way the sfrom a complex negotiation between patient and doctor. The story itself involves
and w atch the slgmfraction slg� ?f nce.. an appreciation and response
m Ica and invokes the bodily portrayal of an event which illustrates the patient's
achieves its local inte difficulties and to which the doctor responds with an appreciation and a
the doctor.
1 . Transcript 3 and Figure 1 4.2
diagnostic assessment. In this way the doctor's professional assessment serves
Frag ment to discourage perhaps any further elaboration by the patient concerning her
experience of her problems. But even before the patient begins to describe
�� n �p down up down up ��
the events, she is sensitive to the concurrent conduct of the doctor and appears
up dow to make various attempts to encourage him to transform the way in which
ay up I felt, he is participating in the talk. Indeed, it may even be the case that the very
P: I
� was com 'I ng up the
�
steps \i:ke tt1is all the w
enactment of the events arises in the light of the patient facing a potentially
recalcitrant recipient; a recipient, who is looking at a note pad and writing as
nods &
/\
/\ /\
/\

turns to turns to she begins to tell of her 'awful' suffering.

Dr: writes P's legs smiles The patient's actions are highly complex and contingent. They not only
prescription P's face
serve to illustrate the difficulties she once experienced, but simultaneously
The patient's success in perf encouraging establish an audience for their performance and thereby achieve the sense
orman.ce and sequential significance of the story. The patient's bodily conduct creates
the doctor to watch thethe sequentIal different, but interrelated, sequential trajectories. It forms an integral part
and thereby achieve rives from t�e of the story to which the doctor responds following its completion. It also, as
relevance of the story de s her bodIly part of the story, encourages the doctor to transform the ways in which he
ways in whi ch she design ins to step is participating in the talk and in particular to watch the performance. In a
comportment . As she . beg .
swm?s her sense therefore, we might speak of the bodily comportment as a 'double duty'
u for the second ti sheIn pa tlc
me,
hips towards the doctor. : �lar, activity, sequentially implicating specific actions from the recipient at differ­
towar ds hiS visual ent locations during the course of the movements' articulation (see Heath,
she swings her hips wee n the pre­ 1986). And yet such a characterization hardly catches the complexity of
bet
field, an area midway e. Just as her
fac the activity; not only is the relevance of the story as an illustration as a whole
scription pad and his ks up, �ur�­ contingent on establishing within its course a visually oriented recipient,
hips near the doctor, he loo. The patlent s but also the activity is accomplished in and through both talk and bodily
ent
ing to the face of the pati of the ove all
nt conduct.
movement, a compone s the eon.r�n­ In this way therefore we can begin to disassemble aspects of the social and
demonstration, engenderouragm �
g h�rn interactional organization which feature in the accomplishment of a particu­
tation by the doct or, enc
tem ora nl Y lar event and provide for its character and uniqueness. Even this cursory
to abandon the prescript in whirch he IS,
ion
glance at the fragment reveals the complexity of the participants' activity
and transform the way s ry of the story. and the remarkable resources which are brought to bear in the production and
participating in the delive ent's face, the intelligibility of the story. In particular it shows the emergent and contingent
On turning to the pati her own legs character of the participants' actions and the ways in which they are inter­
doctor finds her lookingHe at
as she utters 'like this'. c pe looks down actionally organized and accomplished. We can see that while the visual
rforman�e aspects of the participants' conduct is not organized on the turn-by-turn basis
and watches her dramati wn. An d, as � e characteristic of talk, none the less the sequential organization is a critical
as she steps up and do t0 cOlnpletIOn feature of the activity's production and intelligibility. The elicitation of the
brings the performance d octor utters doctor's gaze is sequentially responsive to an action by the patient, and the
WI'th 'terribly' and ththe e
'yeh' , 'yes ' and nod s, patl.. ent sue
. '1n-
doctor's reorientation forms the foundation to the sense of the story and
par tlCl ' patIO the sequentially appropriate reply. The methodological resources used by
cessfully transforms the temporan y participants themselves in the activity's production and intelligibility drive
of the doctor and has him
277

276
CHRISTIAN H EATH VISUAL DA TA
ANAL YSING FACE· TO-F
ACE I NTERACTI ON
, w h'ICh their conduct is interactionally
anaIytic attention towards the ways m objects to enable others to ga sen
accompI'IShed within the emerging course of the event. (Hindmarsh and Heath, 200in3),a Mo se of their imaginary function an
d purpose
re
ways in which objects are found, ha generally, these studies have explored
discussed and ho w the talk ndled, manipulated, examined, referenced,
ACTIVITIES I N I NTERACTIO N occasioned features of the materis embedded in, an d constituted through,
, body f vIdeo The growing interest in objectsial environment.
The growmg � ' -based research within ethnomethodology and through a burgeoning body of res and artefacts in interaction has also arisen
conversation analysIs concerned Wl'th the visual as well as vocal aspects of place studies', These studies haveearch which is commonly known as 'work­
human conduct stands in marked contrast to more traditional contribution , IS, nos through a wide-ranging critique a curious provenance, emerging in pa rt
to the understan d'mg 0f non-verba i behaVIOur, ' Non-verbal behavIOur cognitive models of human-com of artificial intelligence and tradition al
longer treated as a dIS' l'mct channe
, 1 0f communI'catI'on, in isolation from taIk these workplace stu die s draw fro puter interaction (Suchman, 198
7), Some of
t
and other aspec s o u f h man mteractlon, ' Rather attention is directed towards analysis and are concerned with m ethnomethodology an d conversation
the various resources , , , that , ,
partIcIpants ' t0 bear on the accomplIs, h ment
b nng , , interaction and participation whthe highly variable and contingent forms of
of social actions and aChVltl�S, GIve ' n the 'situated' character of human actIvIty environments, environments wh ich arise within c'omplex organizational
and the uniqueness o,f parh�ula� ��e�: t , st dies are naturalistic rather than plished through an array of too ich encompass a range of activities accom­
experimental and delIberate y ,a o � p�iori theorizing and the develop­ have focused on 'centres of coordils and technologies. Many of these studies
ment of hypotheses characten�tic of the more variable centred approaches and the news media, and explorednation', in such areas as transport, medicine
found within many, psychol�gIcaI and social-psychological studies of non­ roles and responsibilities coordinathe ways in which personnel with differing
'
verbal commumcatlon, Most Importantiy perhaps, analysis is directed towards located elsewhere when engaged te the actions with each other and those
developing observatIOns ' wh'�ch, charactenze ' aspects of the indigenous reason- So, for example, in their wide-rangin interdependent but interrelated activitie s,
' that p artlclpants themseives use in the production, and
ing and practIces Suchman (1993) and Goodwin ing study of ground operations in an airport
' " 0f part'ICU1ar social actIOns
intelliglblhty ' and actI'vI'ties " So for example, m the which personnel identify particuand Goodwin (1996b) examine the ways in
" IS conc; rned with building , ,
data at hand, the anal�sIs , a charactenzatIOn and develop a coordinated respolar problems an d events through interaction
of the resources on WhICh the p,ar I � nts rely in accomplishment of the London Underground, we havense, In our own studies of operations on the
activity, with regard to the �ays � ��IC� they produce and coordinate their configure activities to enable oth addressed the ways in which personnel
actions wit� each other dun�g I ping course of their articulation, located elsewhere, to have sense ers, both within the same location and those
:
The interactIonal and sequentIal ,o�nd:��o�s of the activity's accomplishment priate activities (Heath and Luff,of action and to produce sequentially appro­
provide an Imp' ort ant resource m th e l' d entI'fication and description of the studies not only de mo nstrate 1996, 20 00 ), These and an array of related
act10
' ns, and the practices and reasomng ' upon which the participants reIy In ' conversation analysis can inform the ways in which ethnometho
dology and
their production and inte11'Igi'bTt I 1 y, material aspects of conduct and the analYSis of visual, vo cal as well as
The analytic commitment to d e�c;1'b ' �g the resources on which participants we can begin to consider hig action, but also point to the ways in which
rely in the accomplishment of S?Cla a,c�IOn and activities has led to a growing participation beyond the dyadhly complex forms of interaction and co ­
interest m' exp I onng
' the ways m whlCh obJ'ects, artefacts and other features formed the ma ins tay of studieic, mutually foc us ed encounter wh ich has
' I enVlfonm
of the physlca ' ent feature m ' COl1duct and interaction, In part, these extensive fieldwork, provides the s of talk. Video recording, coupled with
studies are concerned w �th readd�essl' :f the dematerialized conceptIOn , 0f
the methodological initiativ resources through which we can build Upon
action found within certam ar�as � soc� science and with placing the object conversation analysis, es developed within ethnome
thodology and
in action at the heart of agen a, mo 1'vation of these studies is to expl�re
the ways in which material resources are u sed and constituted within SOCIal
action and to conSl'der how the sense an d s1'gnificance of objects, artefa ct
and the like anse" m and through mteract lOn, ' So, for example, studIes, haves SU M M ARY

explored the ways :' n wh'ch 1 conversat'IOnaI'ISts over the dinner table onen
' _

, t t0
Ethnomethodology and con
and manipulate ob!ects wlthl " �ta1k during the developing course of partIcu , 1ar
through which we can versation analysis provide an an
alytic resource
activities (Goodwm, 1984), , ow d tors use particular artefacts such as be gin
video, The possibility of cap to exploit the opportunities provided through
medical records in the inte�a�tIOn -:V l' �� patients (Greatbatch et al., 1993; Heath of in situ human condu turing aspects of the audible and visual elements
and Luff, 2000), how parhclpants mu,tually constitute features of compl�x researchers with unpreced ct as it arises Within its natural
habitats provides
scenes (Goodwin, 1994), and the ways m wh'ch1 students embody gestures Jn ethnomethodology and ented access to social actions and activities, With
conversation analYSis, the technolo
gy opens up the
278

279
CHRISTIAN HEATH VISUAL DA TA
AN AL YSI NG FACE-T O-F

of developmg . a socl. Ology which takes the visual, material �s ",-,ell ACE INT ERA CTI O N

Pasossibility
vocal aspects f hum.a� m. teractlOn
0
. ly, as a topic for investlgatlOn
' . senous
Ga rfinkel, H. (19 67) Stu die
s in Eth nomethodology.
.
and analysis. Such studIes w ]] ot replace or supersede the extraordman . '1y
Hall.
Goffman, E. (19 59) The Presen
Englewood Cliffs, NJ: Pre
ntice

rich body of work concerne� :ith talk in interaction, but further enhan �e Goffman, E. (J 963) Behaviour
tation of Self in Everyday.
Ne w York: Doubleday An
chor.
our understanding of how the body and PhYSlCa .
1 objects feature with talk In Gatherings. Ne w York: Fre
in Pu blic Places: Notes
e Press.
on the Social Organization
of
the produ ction and intelligibili : y of everyd aYt�cial actions and activities. In Goffman, E. (1967) interactio
n Ritual. New York: Doub
order to deve1op a SOCIO, .
logy whIch ad.dresses t detail of the body and objects
1
Goffman, E. (19 81) Forms of Tal leday.
in interaction, .It. IS. Im Goodwin, e. (19 79) An intera
k. Oxford: Basil Blackwell.
ortant that we aband n some of the trad't'I lOna
0
. 1
distinctions whIch have .�mfor�ed our
ctional construction of turn
understanding of non-verbal be h avlOur In G. Psathas (cd .) Everyd
ay Lan gu age: Studies in
at talk in natural conversat
ion .
and consider the way� m whIch Vlsua 1, rnat en' al and vocal conduct features
. Irvington. Eth nomethodology. Ne w
Goodwin, e. (19 80) Restarts,
Yo rk:
in the in situ accomplIshment of P articular . . . . In
social actions and actlVlhe pa use s and the achievem
. 0
this way we can begl� t move away from the psychological and cogmhve
. _

�. beginning. Sociological Inquiry

Goodwin, e. (19 81) Conversat , 50: 2 72-302.
ent of mutual gaze at tur
n­
characterization of vIsual �onduct. and the use of artefacts, and b egm t . 0
ional Organisat ion : interactio

develop inslg. hts m. to the socIal. and mterac

London: Academic Press. n between a Speaker and He
10
t' nal organization of activities Goodwin, e. (1984) Notes on
. I, SI' tuaand
' arer.

the range of resources on. WhlCh artici ants rely in their practlca .ted I.M . Atkinson and J.e . He
Story Structure and the Or
ganisa tion of Participation.
accomplishment. By turmng anaF� lc aK�ntion towards the methodologIcal
In
l'
ritage (ed s) Str uct ures of
Cambridge University Pre Social Ac tio n. Cambridge:
Goodwin, e. (1987) Forgetful
ss.
and reflexive found ations . of SOCIaI actlOn, we can develop a distinctive ness as an interactive resou
understanding of the orgamzatlOn ' f hurnan activity in all its intell'lgence and
0
50 (2): 115--30.
Goodwin, e. (1994) Profes
rce. Social Psychology Quarte
rly,
momentary glory. Goodwin, e. (1996) Transp
sional Vision. A merican An
thropo
arent Vision. In E. Ochs, E.A logist, 96 (3) : 606-33.
(eds) Interaction and Gramm . Schegloff and S. Thompso
Goodwin, e. and Goodwi
ar. Cambridge: Cambridge n
University Press. pp. 370-40
n, M .H . (19 96a ) Formulat 4.
Activity. In D. Middleton ing Planes: Seeing as a Sit
and Y. Engestrom (eds) uated
Work: Dis trib uted Co Cognition and Co mm unicat
Press.
gnition in the Workplace.
Cambridge: Cambridge Un ion at
Goodwin, e. and Goodwin
iversity
, M. H. (19 96b ) Seeing as a
Planes. In Y. Engestrom and Situated Activity: Formulati
D. Middleton (eds) Cogni ng
Work. Cambridge: Camb tion and Co mm un ication
ridge University Press. pp at
Goodwin, M.H. (1980) Pro . 61-95.
cesses of mutual monitori
of description sequences. Soc ng implicated in the produ
iological Inquiry, 50: 303-1 ction
He ath , e.e . and Ca mp
Greatbatch, D., Luff, P., 7.
Communication and Huma ion , P. (19 93) Interpers
n-C onal
computers in medical consul omputer Interaction: an examination of the use of
Heath, e.e . (1984) Participat tat ion
of
s. Interacting With Computers
ion in the medical consultat , 5 (2) : 193-21 6.
and nonverbal behaviour ion: the coordination of ver
Heath, e.e. (1986) Body Mo rnal of the Soc iology of Health and Illness, 6 (3): 311 -38
. Jou bal
vement and Speech in Me .
Cambridge University Pre dical intera ctio n. Cambrid
Heath, e.e . (2002) Demonst ss. ge:
rative suffering: the gestur
Journal of Communicatio al (re)embodiment of sympto
Heath, e.e . and Luff,
n, 52 (3): 597-61 7. ms .
R EFEREN C ES
P.K . (19 96) Convergent
multim ed ia technology in activities: collaborative wo
J .M . (1978) Discovering Suicide: the S OCla1 0rganisation
. London Underground Lin rk and
. of Sudden Death. London: Middleton and Y. Engestro
Atkmson, m (eds) Cognition and Co mm e Control Ro om s. In D.
.

Macmillan. Cog nit io n in the Workplac unication at Work: Distribute

Heath, e.e . and Luff,
. e. Cambridge: Cambridge d
· d whistell, R. (1970) Kmeslcs
B Ir . . Con text·' Essays on Body Motion Communicatwn.
m University Press. pp . 94-130
.
,

. P.K . (2000) Technology in .

act ion . Cambridge: Ca
B od en, D . and Zimmerman, D.H. (eds) (1991 ) Talk and Social Structure. Oxford: p0ny
Philadelphia: UniverSIty 0f Ph'ladelphla
I Press. UniverSity Press. mbridge
I Heritage, J.e . (1984) Ga
Heritage, J.e. and Atk
rfinkel and Ethnomethodo
logy. Cambridge: Polity
ethologlque d ans l" etude du comportement
Press. inson, I.M. (1984) Introduct Press.
ion. In ].M. Atkinson and I.e
. Heritage
. . . . (eds) Str uct ures of Soc
Cosnier, J. (1978) SpeClflclte' de l' atftude
1lumam.
I
ial Ac tion : Stu dies in Co
. Psychologic Franqazse, 23 (1).. 19-26. Cambridge University nversa tion al Analysis . Ca
Press. mbridge:

280
281
CHRISTIAN HEATH VISUAL DA TA

. .
Heritage, J.e. and Maynard , D (2004) Practicing MedicIne: TaIIc and Action in Primary
. : Cambrid e University Press.
�
Care Encounters. Cambridge.
Hindmarsh, J. and �eath, e.e. ( 2000) mbodied Reference: a study of delxls m
. . .
workplace interactlOn. Journal of Pragmatics' 32 (12): 1855-78.
. .
Hindmarsh, J. and Heath, C . C . (2003) Transcendmg the ObJ'ect in Embodied Interachon.
the Body and Identity. London: P a1grave.
. 1(, A . (2002) The T aCl't 0rder of Teamwork: collaborallOn
InJ. Coupland and R. Gwyn (eds) " D Iscourse,
'
Hindmarsh, J. and PI'1�IC
.
an d
.
embodied conduct m anaestheSla. Sociological ' Quarterly, 43 (2): 139-64.

C; ��
Hughes, E.e. (1958) Men and Their Worle. e oe NJ' Free Press.
Kend on, A . (1991) ConductIng Interact IOn.. a rn� of Behaviour in Focussed Encounters.
.
Cambridge: Cambridge Umversl'ty Press '
LeBaron, e. and Streeck, J . (2000) Gestures, lmow1edge and the world. In McNeil, (ed.)
Language and Gesture. Cambridge ' Cambridge University Press. va l i dity in
Luff' P. , Hindmarsh, J. and Heath: C . (ed s) (2000) Worleplace studies: recoverIng work
.
.
practice and informIng sys tem deSign ' Camb n'd ge '. Cambridge University Press. on
M arks, D . (1995) EthnographIc hIm.. from 1'1a berty to Asch and after. American oc cu rri n g so ci al
. . .
Anthropologist, 97 (2): 337-47.
Sacks' H. (1992) Lectures In ConversatIOn, olumes I and II. Oxford: Blackwell.
.

Anssi Perakyla
Sacks, H., Scheg 1 0ff, E . A ' and Jefferson,V G . (1974) . A simplest systematics for the
. .
organization of turn takmg m conversation. Language, 50: 696-735.
Scheflen, A.E . (1 964) The Significance of Posture m C0mmunication Systems. PsychIatry,
. .
27.' 316-31. '
Schegloff, E.A. and Sac1(s, 1J . (1974) Openmg up closmgs.
.
' In R . Turner (ed.)
. .
Ethnomethodology: Selected WntIngs. H a rmondsworth: Penguin. As Kirk and Miller (1986: 1
of reliability and validity 1)areanim d Silverman (2001 : 219) point ou
S treeck , J . (1988) The Significance of Gesture.. H ow I' t is Established. In Papers In .
t, the issues
Pragmatics, 2 (1 /2): 60-83. and credibility of (social scientific) portant, because in them the objectivity
Streeck, J . (1996) How to do thmgs With th 'mgs.. 0b}'ects trouves and symbolization. Human
. .
is to produce descriptions of a socresearch is at stake. The aim of social science
Studies, 19: 365-84.
.
' . the Problem of Human Machine Interaction.
descriptions that in som e contro ial world - not just any descriptions, but
� that is being described. Even thougllable way correspond to the social world
Suchman, L . (1 987) Plans and Situated ActIOns.
.
Cambridge: Cambridge um."ers ' t P ess
�� c ou�tability: On Lizards and Aeroplanes. In perspective and therefore repres h all descriptions are bound to a particular
�
Suchman, L. (1993) Technologies
G. Button (ed .) Technology and � e Wor) l.ng Order London' Routledge. pp. 1 13-26. (Hammersley, 1992), it is possible ent the reality rather than reproduce it
vom Lehn, D., H e� tI1, .C .C . and Hmdmarsh, J . (200 i ) Exhibiti ng Interaction: Conduct can be subjected to empirical testinto describe social interaction in ways that
C

' 24 There is no single, coherent set g.

.
and CollaboratlOn m Museums and Galleries Interaction. Symbollc InteractlOn,
(2): 189-217 . analysis of texts, talk and interacti of 'qualitative methods' applicable in all
Whalen J. (1995) Expert Systems vs. Systems for Ex� erts: Computer-Aided Dispatch
. Rea1- sets of methods: different ways on. Rather, there are a number of different
�: and the use of symbols. In so far of recording and analysing human activity
' .
as a Support Sy�tem m o ld Envlronments. I n P . Thomas (ed.) The Social and
interactional DimenSIOns 0f u an computer Interfaces. Ca mbridge: Cambn'dge status different from mere com as these various methods claim an epistemic
_
University Press. pp. 161-83. than the research subjects' own de mon sense, in so far they claim to
report more
of objectivity is relevant for all scriptions of their circumstances, the question
In research practice, enhancithe ng
se methods.
involves efforts to assure the objectivity is a very concrete act
acc ura cy and inclusiveness of recording ivity. It
the research is based on as we s that
lytic claims that are being madell as efforts to test the truthfulness of the ana­
however, take different sha about those recordings. These concrete efforts,
the research is based. Qu pes according to the type of recordings on which
field notes, for example, areestions that arise in the context of ethnographic
of written texts. Field notes different from questions that arise in the context
issues or, alternatively, the can be produced so as to be focused on particular
y can be produced so as to include
as wide a range of
282
 �
RElIABILITY AND VALIDITY IN RESEARCH BASED ON NATURAllY OCCURRING SOCIAl INTER CTlON

'" VALIDITY . �f all conversation anal tic stud'Ies (both on rdmary. conversation
ANSS I PERAKY LA
3: 1 50-1 ). In theotanaly sis The a.lm ; . �
Ha m m ersle y a nd Atk ins on, 198
sible ( question does not arise: the researcher cann rse, co ntrol and on InstItutional interaction) i to pr0duce descnptlOns of recurrent pat-.
events as pos text s such a that is used a s data. (But the researcher can, of cou tern� of social interaction and la.nguage use. Conversation analysis (CA) IS
. . grounding
te n
of writ f a giv text he or she uses .) partIcularly rigorous in l
' ts reqUIrement of an empmcal for any
the focus orange ofentex based . . to be accepted as valid
descnptIOns ( ' I n thOIS respect, CA differs from some
select the pter will dealts that wi th iss ue s of reli ability and valipart dity in research
in con ver­ other forms of discourse analrsi; F::. .�10ugh' .1999; Parker, 1992) and social
This cha video recordings and transcripts, and, in icular,of qualitative constructionism (Gergen, 199 ) Whi ,lze more th� 'openness' of any
on audio or sis. I will focus this discussion on one specifihe
o n analy d out , t
c type s of reli­
question The langu.age �se to different interpretation::���ence underlme . morethat
the actl've
sati only mainly because, as was just pointe
ch qual itati ve methods.search contnbutIOn of the researcher ' ,constructi
111 . ng' tl1e d escnptions he or
re sear ke a diffe rent for m in diffe rent she produces about language use ' N �vertheless, .these issues are complex;
ability and val idity ta on thi s spe cific vari ant of qualitativseofrevalidity ther forms of discourse anal sis an notably. dIscursive psychology (see
second reathat
the fact
fo cu sing
son for re no accessible discussions available on issue however,
there a rsation analytic studies.1 This does not mean, tion l�otter, this volume) draw dire�ly on conversatIOn analytic
,
thinking.
is onve d reliability have been addressed in conversa
and reliabilitnysinofcvali ty an they are addressed more there than in many
that questiosearch pradictice. n fact But what has been lacking is a general
analytic rees of qualitativeIrese
other typ nted discussion abarch. out vali dit y a versation
nd reli abilityainconcontribu tion in �(irk and Miller define reliabilitY as 'the degree to whIch . the finding is
student-oriesearch . The purpose of this chapter is to make mdependent of accidental circum. stances of the research' ( 1986: 20). In ethno-
analytic reion. type of
a specific here
graph'IC research, the reliability of resea ch results entails 'whether or not (or
that directugh the discussion in this chapter focuses on under what conditions) the eth���rap�er wou}d expect to obtain the same
Altho esearch (conversation analysis), the basic issuesore,raireaders wareho se d finding if he or she tried again i e s.ame way (1986: 69).
qualitative rthe context of any qualitative method. Theref d to treat this In the context of ethno ra h �I� rman (�001: 227-8) also points out
relevant in arily interested in conversation analysis are encoudrage dressed checking the reliability is �lo!:lrr:: � �0 assunng the quality of field note�
are not primn example of the kinds of considerations that neestions andadan to be ers and guaranteeing the public access t�_h�r�ocess of the�r production (d. also
chapter as aitative researcher. Even though the specific que alitative methswods, Atkinson, 1983: 14 6 . n c�nversatlOn analytic research
Hamm.ersleyandandtranscripts
by any qual validity and reliability are different in other qu recordmgs are the 'c � atenal' c mpacablc to ethnog,"�
concerningconcerns are the same. primarily co992;n- phe,,' field notes. Accmdingly, the � i;"� of recordI��gs and transcripts has
the basic umentation presented in this chapter concerns d Heritage, 1 to Important Implications for the relia�ny o conversatIOn analytic research.
1
The arg ic research on institutional interaction (Drew an con tribution
versation aSorjnalyt ). As Joh n Her itage points out inatiohisn analy
D rew and e, there are two different kinds of convers eraction as sisan going
o ne n, 19 97
ntity :C
Concerns o f reliabilit as
: reason for working with
this volum'The first examines the social institution of int al institutionse[su recordings and tran riP t

on today: right; the second studies the management of soci(p. 223). The first typech
in its own tion, classroom, medicine, etc.] in interaction' nary conversation':
as corporasation analysis focuses on what is called ' ordi like. The latter one
of conver king among friends, family members, or the clients or amongst
informal talverbal interaction between professionals and
focuses on ls.
professionahodological constraints facing these two typent.s ofIn con
The met partially overlapping and partially differe ysithis.s Hen
analysis ar focus is on the latter type of conversationalong with the ceco,nI­
e anal
Working with audio and :�'de� :�c��dmgs . and transcripts eliminates at one
stroke many of the probl S h ;���apher� have with the unspecified
accuracy of field notes and with the li bhc ac�ess to them. According
to Harvey Sacks, realizing the potential of a��10- or vIdeo-recorded materials
actually gave a crucial impetus to the creatIon of CA:
versation
chapter,
Itwas not from any 1arge interest in Ian u ge or from some theoretical formulation
f what should be studied that I started !it� ta e-recorded conversation, but simply
�ecause I could get my hands on it and. I coulPd study it aga ' and again, and also
consequentIa' 11y, because others could 10�k "t ';;.,h:U :' tudied and make of i;
111

the primary the use of wvritte

issues sucd htheascrit n documems nts
about the rel evance �h8�t ���y could, ii, lomampIe, they wan ed to e���ISagree WIth me. (Sacks,
will discusS dat eria for alida ting clai
a, an of interaction. These issues concern the analysistheofr
versational al contextnot the study of ordinary conversation. Some o alsO . based on them c�n rovlde . for hIghly
. detailed
of an institutal ion , Recordings. and transcnpts
institution I will dition
int erac
scu SS, ho wev er (su ch a s d eviant case analysis), are .
and publIcly accessible representations of SOCIatmteractlOn. Therefore, Kirk
issues thatle to the analysis of ordinary conversations.
applicab 285

284
-

RELIABILITY AND VALIDITY IN RESEARCH BASED ON NATURALLY OCCURRING SOCIAL INTERACTION

ANSSI PERAKYlA VALlDlTV

, on ;hat ' q native research 'issues of reliability ha�e The whole richness of ambulatory interaction can hardly be encapsulated
and � i1le �'s sug gestl , (1 �� : �� �oes no t apply to conv�rsation analytlc using a stationary video camera, say, in one patient room, By collecting
n tIOn e of many ethnographic data along with the audio or video recordings, the researcher
rece Ived lIttl e atte
rt of�Its , h' flca
JuS ' tIon' 0n the basis of bemg fre
cten' sflC 0f 0ther forms of qu
rese arch , CA cl a '
lm s pa alitat '
Ive research, can capture some aspects of ambulatory events. A good example of the fruitful
sho rtco ming s in reliabili ty cha ra combination of ethnographic and audio- or video-recorded data is provided
especially ethnography , by Goodwin (1994, 1995). Moreover, by the use of multiple cameras, recordings
can be made that are both comprehensive and accurate. Multiple cameras
of ",,,,, ,�nr'rHe:!l1I data also need to be used when the interaction involves multiple sites which are
" g m axI"m um
Sec urm ength in terms
Although audio- or vldeo-
, record�d data have intrintosicbestrpaid to the inclu­ connected using technical means, such as monitors or telephones, Charles and
ss, specla1, attention needs specific events (such Marjorie Goodwin, for example, have used multiple cameras in recording
of accuracy and public acce
siveness 0 f such data ' ,
Vide. o or aud10 recordingss or of
pu bli c me etm, gs) maY the work of the crew on an oceanic research vessel (Goodwin, 1995) and the
activities of air-traffic controllers (Goodwin and Goodwin, 1997). Similarly,
edI ' a I on sul tat ion
as telephone conversatIOns,s 0mf SOC�lal m � tion, including (a) medium- and Heath and Luff (2000); Luff and Heath, 2002) hAve analysed convergent
entail a loss 0f ' pact 0f exts
som e a0p ect t e rac t
" , ses{ ) ambulatory events and (c) Im activities in various technological working environments,
proc s
long-span tem
and other 'non
por
-con
al
vers a �l�na �
� , odalities of action, The potential loss can be
n.
prevented with app ro pna e arrangements in the data collectio Documentary realities
Written documents (and their production and use) are important for social
Temporal pro cess
es
, research h �S brou ht a new kind of
temporality into �he life, as a domain of signification of its own, or, as SInith (1974, 1990) puts it,
C onv ersat ion anal Y �lc , cal ana sis ' s� uential organization of intera ctIO,n as a 'textual reality', Written documents also constitute a domain which is
logI
ce ntra l fo cus of s
operates in and through the
ocIO
reiaflVe ' t� 1 ing of act ion s,
As ethnographIc
wo s are, also organ-
rld
in contact with the domain of spoken interaction and which in some events
organizes some aspects of it. As Firth (1995: 205-11) has shown, in international
research has repeatedlytesho wn however, local social his d
longer mpor�l s, P ans (letsalo ne the torical time focuse rx) ,
business communication, written messages (communicated through telex and
ized in terms of , cio log lca 1 wo r k o f Du rkheim' We ber an d Ma , fax) relate in many ways to the organization of the telephone conversations
on in the classlca1 ma cro -so nts
' ls, for exam pIe , managem ent 0f chronically ill or dyda ing, patle
' s
lly actIOn
between the trade partners, Therefore, full understanding of some of the
institutional activities conducted by telephone is not possible without the
In hospIta gh the evo lvi ng
involves complex traj ecto ries sh d� d m' an d thr ou
ss et al., 1985; Sudnow, analysis of the prior written messages that inform their production, Similarly,
Of staff and patients (Glaser an trauss, 1968'' 'Strau 1 n of events such as child in medical settings the content and the ordering of some of the questions that
, SOCl' a1 serVI'ces , the recogm t'o
m
1 967), Simila ,
t
rly,
m vo I ves I ong -sp an pro ces ses WI' th a multitude of agents and the professionals pose to the clients can be strongly influenced by clinical forms
mistreatmen ,
their negotiatIOns at d1' fferen t sites (see Dmg ' wa11 et al" 1983), , that the professionals need to fill in, Thus, even though every question has its
reco � d'l 0 f '
l ng le en cou nte rs, there is a nSl( 1
local interactional management which can be observed in the recording, the
In the research bas edr-ter
e lo nge
on
m te P � � : l P r � cesses will be lost from signta h,t. To logic of the questioning as a whole may not be derivable from the vocal events
that some of ,
thes
m h app enm ' g , lon glt, udma ' I study designs can be used, He ord , ges
mg
only, Therefore, it is important that the conversation analyst carefully collects
prevent thIS fro based on rec and uses all the relevant written documents, along with the collection of
and Lindstrom, (19"98), by for exam P1 �, ,re Port research o had recently g,iven
a hea I th VIS Itor t0 a m othe r wh
greSSIVe1y
recordings (for further examples of this, see Maynard, 1996; Whalen, 1995),
of six co nsec
ir
utlv
anal
e �
ysIs
lSI t
foc
s
uses � n the w � � s in which the mother prothe health
birth, The
es m ora lly prob lem atlc m � tena , and the way s in which Different aspects of social organization
disclos res, In sum, by appropriate research design, conversation analytic studies of
visitor manages these disclosu institutional interaction can be made more inclusive in terms of different layers
nts in of the organization of action, However, it also needs to be pointed out that
an's terms, the participa conversation analytic studies do not aim at describing all aspects ofsocial organization.
nts
Ambulatory eve Go Hm
abo ut , dom
m ' g thi n g s In
: U 1ar ent1't'les, that is human ambulatory
People move , are 've hlc '
e mte rac f Ion v�r (This is, of course, true concerning any other methodology as well,) The
any face-to-fac a participant or as an obpser
units' (1983: 7), � o r an
g, th
� one
IS n
wh
ust
: � �bviOUS The ward round, for exam , IS,r­
s act ed as le organization of verbal interaction in face-to-face encounters and telephone
conversations is the domain in which adequate conversation analytic studies
in a hospital set tm : , W, l' ' 1e event with a number of alte
from the p rofe ssio nals ' � om t of l
��
:r lpan .� �Ys event moves about in patients'
s,
can rightly claim superior reliability, and this is indeed the home base of CA
grou ps of patlen t pa rtlC methOdology, In studies that focus primarily on other aspects of social
nating sub ward,
rooms and in the corridors of the
287

286
 2

ANSSI PERAKYlA VALIDITY REL IABI LITY AND VAL IDIT Y

I N RES EAR CH BAS ED ON NAT
URA LLY OCC URR ING SOC IAL
INTE RAC TIO N
organization (such as textual, pictorial or technological realities) other methods the sel�ction of wh at is analysed in det
may be more suitable. transcnpts only (cf. ten Have, 199 ail is usuallY done on :he �asl. s of the
9: 77- 8) . The qualIt. of transcnpts m research
on naturally occurring interactIOn " seems to vary ygre atly . Not· onIy are the
detal'1s of intonation and pro
Improving the reliability of CA in its own field
probl ematIc. , Whole utterancessod Y som etI. mes omItte . d bu .
(es . 11 ' . party' SIt. t' what IS mo re
The claim of superior reliability in studies of face-to-face interaction needs to missing from transcripts in die��� � n m ltI-
be justified, however, in each single piece of conversation analytic research. adequately designed and constu �
:t t erW�Ise have beenuaser tions) can be
iously an d
The key aspects of reliability involve selection of what is recorded, the technical ducted .
Transcription is a skill that can
quality of recordings and the adequacy of transcripts. training. It is extremely useful if an oni b acqUIr. ed th�ough long enoug
beginner. This is most easily do ne bexf:ri:nc ed trans�nber can supervise a
h
Basic selection ofwhat is recorded arises, of course, from the research problem.
But after this has been done (i.e. when the researcher has decided to record some of the beginner's transcripts . In lac the more ex� enen ced one correcting
encounters in a specific setting such as classroom, doctor's surgery, edu­ for anybody preparing them. another , the correctIOn of transcripts is use ful
cational counsellor's office, or the like) the researcher still has to make some the th·mgs that one has not not' Ice resear. onche r can alwa ys 1lear some of
very consequential choices. The most important choice is how much to record.
c
' d . C orr ech by
culture of shared practices in measurin
g
.
pau ses , . c61ona
mt
1ea� .ues a1so enhances a
There is a limit to how much data a single researcher or a research team .
It is advisable ta mc I d e ma ny aspect tlOn, and so on.
can transcribe and analyse. But on the other hand, a large database has definite transcripts (for the conversation anal '
u
s of v oc� I exp ' .n 111
reSSIO . the initial
t' t cn. phon con entions dev
advantages. As the analysis of data in conversation analytic studies usually
progresses inductively, the researcher normally does not know at the outset
by Gail Jefferson, see the Appendix
198 4: ix-xvi). A ric h transcript is a tres �� ����
[t s me and tkmson and Heelo � � pe d
rita ge,
of the research what exactly the phenomena are that he or she is going to focus cn'b'mg, the researcher cannot kn' ow wh ource of analYSIS ' at the tIm " e af trans-
.'
on. Therefore, it may turn out that he or she wants to analyse events that do important for the analysis . After the ich of the detaIls wIl. l turn out to be
not occur very many times in each single recording. For example, delivery of results are published, however, somanaIYSIS' has been accomplished and the
the diagnosis in medical consultations is such an event: there are consultations the analYSis can be left out. 'SimplifIed' e of the . ptsspe cia. l no ta t'IOn not used in
where no diagnosis is delivered, and in many consultations, it is done only anaIYSIS' eaSler,
·
tran scn can make tl1e reception . 0f
especially if the aud 1en
" ce IS not speCIa the
once (d. Heath, 1992; Perakyla, 1998). In order to be able to achieve a position .
In sum, the relI" abIlIty of observations " lIze d in CA .
where he or she can observe the variation of the phenomenon (such as the delivery in any other empirical method) :an an in conv:rsatron analytic research (as
.
of the diagnosis) in any reliable way, the researcher needs a large enough The method itself does not gu n t Iy b� a�l�leved through serious effort .
�
collection of cases. Therefore, he or she may need to have access to a relatively stud'Ies, proper attention needs to be pee aid
relIab1l1ty In conve rsa fIOn analytI
' . 'c
large database. In practice, a large portion of the data can be kept as a resource of recordings as well as to the ad equacy tO the selectI�n and technical quality
that is used only when the analysis has progressed so far that the phenomena af th e transcnpts.
under study have been specified. At that later stage, short sections from
the data in reserve can be transcribed, and, thereby, the full variation of the VALI DITY I N CO NV
ER SA TIO N A NA LY
phenomenon can be observed. TIC

The technical quality of recordings is a decisive issue: if something is lost from The validity of research con ns th . . of obs
sight or remains inaudible in the recording, there is no way of recovering it. or not 'the researcher is caBcer e l�terpretatIOn
in wh a � �e asu red
ervations: whether
It may be extremely frustrating to have some badly recorded sections of events
that at a later stage of the research turn out to be of primary importance for
and Miller, 198 6: 69; d. Altheid! an
In the discussions about validit d / b
0 ns n, 1 94; SIlver
? �the
r right name' (Kirk
man, 200 1: 232-48).
the analysis. This kind of frustration can be minimized already at the planning research, there is an under
between the 'raw' observatlying
c ly��
b:C:k !o: assumptI�on about
es m the onte xt of quantitative
stage of the research by paying enough attention both to the quality of the
equipment and to the arrangements of recording. The crucial aspects of quality for or represent. Responses to �
ions and t e �. ssues that these observ a separation
questi'onnalres for example, can be atio ns stand
include the sound (quality and location of microphones) and the inclusiveness VaI'd
1 representations of und er I ym g SOCIa 1 phe
. . ' mo re or less
of the video picture (the location and the type of the lens of the camera(s)) (d. d �nts' attitUdes or
values (d. Alkula et aI ., 199no).meAn�, �uch as the respon-
��
Goodwin, 1992; ten Have, 1999: 71-3) . The recording technology is developing �hIS kind of approach the cor IS .m stark contrast to
quickly; digital recordings offer many advantages in comparison with older Interaction, not as 'a scr:een e of ItS
' very aIm IS to mv
. ted estigate talk-in­
on which are prOjec :her
� ; � � �;
Schef,loff, l 92a XVl?ll). ThIS. com
analogue technology. phenomenon in its own ri ht p roc esses', but as a
The adequacy of transcripts is equally important: even though in a proper naturalistic description of t :
mitment to
analysis of data the recording needs to be listened to and watched, at least action taking place within e i te ac It o er (Goffman, 1983) and the social
that or er c . a so Sacks, 1984) gives
a distinctive
288
289
 RELIABILITY AND VALIDITY IN RESEARCH BASED ON NATURAllY OCCURRING SOCIAL INTERACTION

AN SSI PERAKY LA VALIDITY

e the transparency ofanalytic .

But just as in Wittgenstein's philosophy, 'a1th ug� the result . . IS. simple,
shape to the issue s of validation in CA. These includ
xt tur n ' , deviant case analysis,
the
questions abo ut �
.Its method cannot be if it is to arrive at t at re�ult� �WIttgenstein, 1975: 52). In
n thr oug h 'ne versation analytic
.
cla im s, val ida tio
n, the generalizability of con
CA the complexities of the method mvo ve ot er kmds of issues of validation.
cha rac ter of int era ctio
ins titu tio nal
stical techniques.
findings and the use of stati Validation through 'next tum'

As Sacks et .al . p0 inted out, research on talk-i n-'mteractIOn

. has an mherent
.
The
of
t philosophy, methodologIcal resource that res h writte� texts lac�s: 'Regularly . . . a
icus, turn's talk will display its s eake:�;�n��rst�ndI,ng of a pnor turn's talk, and
Tractatus Logico-Philosoph
Wittgenstein pointed out tha rification
In set of pro po sitions but an activity, the cla whatever other talk it mark� itself S d'� cte to (1974 : 728). In other words,
in the unfolding of the interactio:
t a
th �ntera �tants dIsplay to one another
un de rst oo d, is no thi s activity
rightly
hil oso ph ica l pro po siti on s abo ut the world. The method of res ults of
of non-p the
r thinking are complex, but d of paradox . '
ots ' in ou their interpretations of what is ��. ����:�;�lally of wh� t was going on in
th? immediately preceding tur:�� l
is com ple x be cau se the 'kn ilar kin
philosophy are simple (seeme Kenny, 1973: 18, 101-2). Aofsim ults is characteristic ge and Atkmson, 1984). From
1 atlOn
of tho d an d the sim pli city res thIS fact .arises a fundamental val'd ' procedure 11lat· IS
. used in all
between the com ple xit y .
sitive conversatIOn analytic research:
of CA, too. analytic research exhibit, inceayopou have
idity:
n
ller (1986: 22) called apparent val
The res ult s of (go od ) con ve rsa tio on .
manner, what Kirk and Mivin ced tha t the y are tra nsp are ntly true. A con­ But while understandings of other turn' s talk are dIsplayed to co-participants'
they are available as well to pr0feSSlOna1 eanalysts' , wh� are thereby afforded a
.
the m, yo u are con Pomerantz
read
tio na l act ivi ty cal led 'fis hin g' may serve as an example. Anita ve rsation can
. ,
proof cnterion . . . for the analy sis of w hat a turn s talk IS occupied with. (Sacks
versa
a pa pe r pu bli she d in 198 0 how participants in a con at they et al., 1974: 729)
showe d
ctl
in
y 'fis h' for inf orm ati on fro m one another by tellin cerwh g
's 'lim ited .
indire ents displaying their pro du
thems elv es kn ow . De scr ipt ion s of ev the oth er par ty At the beginning of this chapter, it was . P ointed out that CA dIffers from those
' to the rel ev an t fac ts ma y wo rk as a device for invitings (assuming, of . . expressions. Now we can the
forms of discourse analysI's and. SOCIa1 construction' Ism w IC
h' h emphasize
access version of the same issue open-ended ness of the meaning of alI I'm�Ulstic see
to disclose his or her authoisrizineda position of having privileged access to the the reason for this: even thoug h the meanmg of any ex , . 1'f considered
preSSIOn,
course, that the other partymics are at work in cases like the following: . .
In ISO . is extremely open-e d d y utteran e that is produced in talk­
1atIon,
in-interaction will be locall inte� r:t���; the par�lClpants
relevant facts) . Such dyna �.
of that interaction.
(1 ) 1 B: Hello : : . In the first place, their inte�pretaKon is d splayed the next actions after the
111

2 A: � utterance. Hence any interpretat 'IOns tIlat conversat'IOn ana1ysts may suggest
3 B: Oh :hi: : 'ow are you Ag
ne: :s,
y.
can be subJ'ected' to the 'pr00f proced ure' outl'med by Sacks et al.: the next
bee n bus
-1 4 A: Fi:n e. Ye r l ine 's
.hh my fat her 's wif e cal led me
turn will show whether the interactants,th.emselves treat the utterance in ways
fu (hh )-
5 B: Ye uh my
ls me : : , .hh I can alw ays talk that are in accordance
. Extractwith the ana1yst s mterpretation .
6 . . hh So wh en she cal Therefore m 1 shown above the utter� nce produced by B in lines
c'n afford it'n I can 't.
7 fer a lon g tim e. Cu z she 5-8 provides a proof procedure for the p m
" terpretatIon suggested by Pomerantz
8 hh hh he h .eh hhh hh
(Po me ran tz, 1 98 0: 1 95)
4)
concerning A's turn in line 4 . (What omerantz suggested. was
side'. (what A did in line can operate as a 'fish'mg d eV.lce,, whIch
that 'telling my
. indirectly
limited access to relevant facingts .
e1ICItS
In Extract 1 abo ve, the des cri pti on bas ed on
ow) works as what Pomeran tz called 'a fish
v:e see, Pomerantz's interpretation P a
�!
an authoritative version of the :::� � rom t�e I.nterlocutor.) And as
test: m Imes 5--8, B gives her
given by A (marked with aningarrB's ort in the ne xt turn. By telling
en busy, A makes it relevant for
eli cit ins ide r's rep fIrst-hand account of what had happened.
. In much everyday conversation anal btic w0:k, th'lr:gS are �ot as nice and
device', succes sfu lly
ns ab ou t the lin e ha vin g be
her observatio talking . SImple as in Extract 1: the next turns maY e ambIguous m relatIon to the action
B to disclose to whom she was ivi ty lik e 'fis hin g' tends to 'ring a bell' as soonody
as
performed in the preceding t�: H�wev.er, �he 'proof procedure' provided
The descript ion of an act in which everyb
out it. 'Fishing' is something by the next turn remains the p i 'or�lal cnten �n of validity that must be used
anyone stops to think abent rol es. Bu t un til Pomeran s article, this activit
tz' y
as much as possible in all conversatIOn analytIc work.
ted in dif fer ver y
has participa
des cri bed for ma lly . Th e res ult s of Pomerantz's analysis are(1986)
had not been transparently true, or, in Kirk an
d Miller's
simple. Her argument is nt validity'.
terms, it has a genuine 'appare
291

290
 RELIABILITY AND VALIDITY IN RESEARCH BASED ON NATURAllY OCCURRING SOCIAL INTERACTION

AN!) !)I PERA KYLA

VALIDITY
the participants discloses that they, too, treat the case as one involving a
departure from the expected course of events. ]f the deviant cases show
this kind of property, they provide additional support for the analyst's
analysis .
Devia n t case . t 1rns of talk conversatIOn
ceSS Ive 5). The
' teraction (Hentage, 199
. . Ons . b et we en suc '
By exammmg the rel atl'mg er.ns 0f m
1

regu lar patt

.

. initial claim that the regularities found in the first phase of the data
at est abl '
IS l1 ' . ns betw een,
analysts aim een actI Ons ( such as the relatio of , fIs. hmg analysis 'are methodically produced and oriented to by the participants
cer
P atterns con and 'gIvn re1 a f l n
� � b. etw
.
ort' in the case , ' as normative organizations of action' (Heritage, 1988: 131).
'telling my SI'de' Ing. an aut110n'tative reper n, the analyst s next task, Extract 2 above is an example of this type of deviant case. After A has
ha vmg stabli she da patt
described abo ve). Aft
ex
er
am . e· deViant. cases.. cases where 'things go differently
m
�
. failed to respond to B's initial 'fishing' turn by an authorized report of
is to search for an d
e 1 eme n t o f the sugg este d pattern IS not the events, B asks directly to whom A had been talking (arrow 3). Through
_ most typically , cas es where an
ed elements . her question, she openly requests the information which the fishing device
embles the technique 0f ' ana
. expect
associated WIt he other SIS
h t . .
I n CA cl ose l y res .
1yt'lC (arrow 1), according to Pomerantz's analysis, solicited indirectly. This
case ana 1Y Atk mson,
The deviant
· eth � graphIC studIes
m
. ' (see Llammersley and e cases
1
shift to open information seeking after an unsuccessful 'fishing' attempt
induction' often used n,
1983: 201-4; Sil 198 5. � 111 -1 5,' 2001 : 237-8). For the analyst, thos indirectly confirms B's initial orientation to the 'fishing' as a device which
verma pattern are deviant. Ratherfthan can be used in indirect solicitation of information.
c

1
. ruc t e d
.
that .do not fIt the m d uc f vely c onst
cases , tIle analys, t is' encou
raged to ocus (2) Clayman and Maynard (1994) point out, however, that there are also
ese disc re pa nt
puttmg aSI' de th deviant cases that cannot be integrated within the analysts' construction
particular attention on them.on ,fIsh . mg . ' , Pomerantz (1980: 186-7) presents a of the participants' orientations that normally produce the regular cases.
In her well- kno wn paper . n 0f events dl'splaying its producer's. 'limited
. ptIO In dealing with these cases, the analyst may need to change his or her
.
deviant case m wh lC ' h a d e scn
0 se her autho riz
ed verSIOn 0f the construction of the participants' orientations. A classical example is
access' does not lea
'
ISC 1
d the oth er party to d Schegloff's (1968) analysis of a single deviant case in his corpus of 500
event: telephone call openings. In this single case, unlike the other 499, the caller
spoke first. The analysis of that single case led Schegloff to abandon
me mill us?= .
1 A: . . , dju j'see his initial hypothesis (according to which there is a norm obligating
B: all day. en the line
(2) tryi ng you
=.hh h No:. I wz
2 the answerer to speak first) and to reconceptualize the very first moves
�II, hh I m 9 nna
s
3 wz busy fer like hour "
oh: :::: .hhh hhh of telephone calls in terms of the adjacency pair 'summons (telephone
��� l
4 A: h h · · · · ·
r i n a ittle whil e help
yer broth er ou:t ringing)-answer'. In the deviant case, the answerer did not produce
l:d
5
B' G oo
the relevant second pair part, and, accordingly, the caller reissued the
� (3c;; �
6 s some he". . l p ,
hh Cuz I know he need summons by speaking first.
(3) There are also, however, deviant cases which cannot be integrated either
7A
8 ((m ournfully» ,
9 B: . .
hh Y e.ah . Yeh he'd -
menti on' that tlhd ay.= into the existing or into a reconceptualized hypothesis concerning the
1 0A ' = M-hm ,= , '
\;;;: t participants' orientations (Clayman and Maynard, 1994). In these cases,
B� ��! �;tZ� 1 980: 1 86·-7)
Who wlh Ik'
11 .hhh Uh: m, .tlk .hh h an explanation can be sought from the individual contingencies of the
single case. Normative orientations or strategic considerations other
. e ce about A's line having been busy than those that usually inform the production of the pattern may be
In Extract 2 ab ove, B repo ts her e���:{ :ttern identified by Pomerantz, this
per invoked by the participants in single cases, and these other orientations
(arrow 1). In term the
�
mte ; actl ls of
s
uld
of
m ake re e; a s� bSe quent disclosure of the detaiab ove,
or considerations may explain the deviance.
kind of telling sho eab le party. In the extractsequent In
r, mo re no W 1 e d g
n. Instead, A sh1·fts the topiana
the event by the othe c in her sub sum, deviant case analysis constitutes a central resource for testing
ppe
ho weve r , this do es not
0
ha
ra me wor k of the lysis of 'fishing'an, hypotheses in conversation analytic work. Therefore, the researcher should
turn (arrow 2). Th ct 2 as a d�vl. ant case In a more recent paper, Claym es
ere f re , with in the f consider the deviant cases not a nuisance, but a treasure. The meticulous
we can consider E xtra ve outlIned th ree �rff at deviant cas analysis of those cases gives impetus, strength and rigour to the development
1 ere nt w ays th
4) of the analytic arguments.
and M aynard (199 ha
can be dealt with :
ctants'
me s de via nt es C ��1 b .
show n to exhibit the intera
(1) Someti on to the sa�e consl e;��10ns and normative orie condUs cttha0
cas nta ti on �
orientati ose cases, something in the
produce the 'regular cases. I n 293

292
 RElIABIlITY AND VALIDITY IN RESEARCH BASED ON NATURAllY OCCURRING SOCIAL INTERACTION

AN!!S I PERAKYLA
VALIOITY
be primarily connected to the organization and dynamics of talk which can be
hara cter
con .
cernmg the institutional c even better understood without reference to the 'institutional context',
of claim s
Validity
, n 0f varId ' tY Another key issue addressed by Schegloff (1991, 1992b) involves what he
of in tera ction
, ' v e rese arch a cen tral dim enS IO J
calls procedural consequentiality of context, He argues that it is not sufficient to
In both qualitative and den quanbtatl ' obser-
resp on . ce bet wee n a th eoretical paradigm and 0the f ten used
say that a particular context is oriented to 'in general' by the participants
involves the cor nstr ue t val'dI 1 ' ty' is a term that IS ' in interaction, but, instead, it has to be shown how specifiable aspects of the
th e r sea rc he r 'Co 198 6: 22) ,
vations made by �
es an d
'Zeller 1 979: 22-6; Kirk and Miller, , context are consequential for specifiable aspects of the interaction, The goal
mm
in this context (Car ns between th eore ' tIC erva tIOns
re l '
atIO ' al concepts and the obsabove, t11e is to make 'a direct "procedural" connection between the context , , , and what
in volv es the pointed out actually happens in the talk' (Schegloff, 1991: 1 7). What is said, when it is
It
that are supposed, to reprCA esent those concepts. As was " nslfles '
I'IStic des cription de-
natu ra . mte said, and how, and by whom, and to whom, may invoke the context; the goal
t h a t plac es. on . r, t he
Primary emphaSISany ordinary concerns 0f construct validity, Howeve , of the conversation analytic research is to explicate exactly how the things
the relevance 0f m , rch on institutional interact IOn ( see said brought forward the context.
f1on analytIC resea need t0
-

ex pa nsio n of,
con vers a 1992) has rein voke d the Schegloff's emphasis on the procedural consequentiality of the context
H eritage, tIns vo1 bet , Drew an d H en 't age, '
ume '
wee n obse rvatl ' ons an d concepts also in conversatIOn has an important corollary, If a piece of research can pin down specific pro­
cons1' der the relation cedural links between a context and talk-in-interaction, it is likely that these
analytic studies,, alytl , c research on I'nstitutional interaction, a central observations not only are relevant in terms of analysis of detailed organization
In con versatI On an the researcher have for clalm- ' of interaction but also contribute to the understanding of the context per
this , wha t gro un d s d oes
question of vaII' d't1 Y IS' e
'
'
IS focu sm, g on IS " m any way 'connected to' some , a se, Standard social scientific understandings of professional and other con­
ing that the talk he or sh of inte racti on take s place m texts are often based on rough generalizations concerning the professionals'
iec
? The fact that a P r se determine the institutional e tasks, clients' roles and the relations between the two (d, Hak, 1994: 472),
institutional frameworkor
hospital or in an office, � ulaexa
d t pe
:nple, ct � � re w and Heritage, 1 992 : 18-21 ) , Conversation analytic research goes far beyond such generalizations, Thus,
r m tera
character of th1at partIcs and arrangementS Inay or may not be present in anyt for example, the studies of Heath (1992), Maynard (1991a, 1991b, 1992) and
� � �
Institutional ro es,, task, they may or may not be present at particular momen s myself (Perakyla, 1998, 2002) on the delivery of diagnostic news have involved
Particular inte, ractraIOctIO ns, programme not only a detailed description of the specific practices found in medical
ula r mte ' ns, If they are' the conversation analytic and the ana1yst consultations, but also a specification of a central aspect of that context, namely
in partic ce , observable to the participants
IS
presuppos es thei r pre sen the dimensions and character of medical authority,
, These two fundamental concerns of conversation analytic research on
alike , v "
hdi ty of claim s co ncer ning the inStltutI' O�al
" na for the , 1992b), The fIrst institutional interaction constitute a validity test for the claims concerning
by Scheg10ff (1987, 199 1nde
Two basic cnte �
utlm ed
' n ' "r11 ere are i
n o
of categorz,za t10
ta lk h a ve bee hm" te1Y many the institutional character of interaction, 'Relevancy of categorization' and
ncy
of
eva
character re 1
.

, 'procedural consequentiality of context' are something to be demonstrated

, on: we m catego:Ize
criterion co nce rns th e racti ay
by the researcher. In demonstrating them, the researcher will focus on par­
any inte
,b of gen�er, a ge social class, education, occupactIO
ally aV 'labl e for h�n,
aspects of context poten asIs n ticular phenomena in interaction, such as lexical choice, turn design, sequence
on the b ' inte ra
one another
an d so on, an d . we m ay un derstand the setting of ourargue , 'n}e organization and overall structural organization (Drew and Heritage, 1992:
income, race, ntary unfoldin of interaction,
Schegloff � 29-45; Heritage, this volume), Where the workings of context will be found
Ch of
1 ect an� display
accordingly, In the mome in thei r co ndu ct WhI
ly and to get her , se nt, or are in a single piece of research cannot be predicted in advance. This unpre­
P arties, sing ny asp ects 0f context they
are making releva dictability arises from the inductive character of the conversation analytic
the indefIm y
' ' tel -r: 'a , moment' (1987 : 219) ,
Iate professional analyst enterprise; it causes both the fundamental difficulty and the exceptional
invoking, for the Immed oble m f r � 1eva nce' requires the fascination of conversation analytic research,
Awareness of this 'pr � � f 'importing' institutionaout l context
� on , Th ere IS a an er going
to proceed with cau t
y t m a � t�m ted to assume, withcation of a
t ; f tJre of talk is an al
prof eSSI O nal anal �
to data , The
into the details of data, tha �;t
� i �; l a � th �ty'
� or 'pro fess ion
indi
dominance' � Generalizability of conversation analytic findings

A crucial dimension of validity of conversation analytic (and any other)

tex t (s ch as Scheglof
particular con �
mte rac �
IO � Such stip ulation for context m atu ay,
d prem rely, SO
research concerns the generalizability of the research findings (Pomerantz, 1990;
having affecte d the
res lt
�
n t e :�� 1 sis being terminate
ly underst�od, cf, Alasuutari, 1995: 143-57), Owing to their work-intensive character, many
(1 991: 24-5) argu
ere nt
es,
org am
�
zatl
�
� n ' : i � h e talk is not thoroughof
cations the workU1gtos
conversation analytic studies are based on relatively small databases, How
that the inh gmm g maY appe ar as indi Widely can the results, derived from relatively small samples, be generalized?
Phenomena which in theexbt� may�m a more thorough examination turn out
of an 'institutional cont 295

294
 II

RELIABILITY AND VALIDITY IN RESEARCH BASED ON NATURALLY OCCURRING SOCIAL INTERACTION

ANSSI PERAKYLA VALIDITY

typ e of few sites only does not warrant conclusions concerning similarities in the
. r of . problem is clos ely dep endent on . the the
the base- professionals' and their clients' conduct in different settings. In that sense, case
ThI . s cha racte
ana 1 Y flC research . In studles
' , 0f ordinary conv ersa tlO
to
n,
t1le w 1101e studies on institutional interaction have a very restricted generalizability.
con ve rsatl On . IS . that the results are or sh ould be gene rali zable . flC
assu mpt lOn n acr oss l'
mg ms However, the question of generalizability can also be approached from a
different direction. The concept of possibility is a key to this. Social practices
line .
satlOns, and . t a cer tai n ext en t eve
domain of ordinary conver thoug: 1t maZ be that thekin most primordial
tura l b oun darie s. E ven g or adjacency that are possible, i.e. possibilities of language use, are the central objects of all
.
and cul structu res - su h as tur enings 0f te1ephone n-ta
co nver satl Ona 1 pra ctice " a nd
h as op conversation analytic case studies on interaction in particular institutional
unI. versal, there are. oth ers: . ' suc 4' Sche gloff 1986), which
J

Pairs are almostop- Hl 199 settings. The possibility of various practices can be considered ge�eralizable
r:stra.' 199 1'f nt cultures . This variatio'n can only be
L d S t rom "
calls (see Houtko St��eatlO n m dif ere
even if the practices are not actualized in similar ways across different settings.
show considerable van l accu 0 dies on ordinary conversadtio n For example, in my study on AIDS counselling in a London teaching hospital
0 (Perakyla, 1995), the research objects were specific questioning practices used
gra dua m u i atlO ' n f s t U y
tackled through so cJal. ilieux . Bu t let us focus now on the stu f
ltu res n d s. . by the counsellors and their clients. These practices, <;trising from the Milan
�ere �. e robl em is posed in different term
in diffe re nt c � �
institutional mteractlOn, wdIes of lns�tutional interact ion, expl icit com parIsons School Family Systems Theory, include 'circular questioning' (eliciting one
In some (advanced). stu are made. M I'ller a�d Silverman ( 1995), for example, party's description of his or her mind by first asking another party to give his
between different settmgs approach In . descnbing talk about troubles in two or her account of it), 'live open supervision' (asking questions in such a manner
. nire counselling patlents. cuw1har,0
pa ra tive �
. that the delivery of the question is done in two stages, via an intermediary)
applied the com.
Bn . . sh 'haemoph·l1 la · ce
fa�.l1y �he���! c;nt . l� the
counselling settm gs: a b. . United States. In partl and 'hypothetical future-oriented questioning' (questions about the patient's
are HIV -positive a�d �lant :: f discursive practic es in these life in a hypothetical future situation). These very practices were to a large
le e extent developed in the particular hospital that my data were from, and it
mtro�ble ��finitio
they focused on slml � ns, trouble remedies and the
ncer n ed lth ne is possible that they are not used anywhere else exactly in those specific
settings : those co . nt�s trou s In a recent paper, Drew (2003) has goon
�
_

social contexts of mpanso the clIe . ns f i�stitutional practices. He foc uses ways that were analysed in my study (see Perakyla and Silverman (1991) and
even furt her. in co . t or up sh ot of the preceding .talk Silverman ( 1997) for some observations on the wide variety of approaches in
form ulatwns, l.e.
glS
ares the uses of formu1atlOn
. u tt er ance s that prop ose a s AIDS counselling in Britain). Hence my results cannot be directly generalizable
197 9 .
) . D rew cO mp
( d. Heritage. and W ats on, tiations, .radio cal l-in to any other site where AIDS counselling is done.
in four settmgs -- n ews othe intervIews, workplace newgothis practIce. IS. s1laped However, the results of my study can be considered descriptions of
h rapy an d s h ows ho
0 ser,; in its spe
pro grammes an d psy � cific contingencies . questioning techniques that are possible across a wide variety of settings. More
se ttm g, so as t raction
differently in each
datab ases � a�lalyses of institutional inte
specifically, the study involves an effort to describe in detail how these
It is likely th at as the
dies like re � s d Miller and Silvertackles man 's will questioning techniques were made possible: what kind of management
lV::pproach directly differences
gradually acc um ulate , stu the of turn-taking, participation frameworks, turn design, sequence organiza­
re co mm o n. The com par a tion, and so on, was needed in order for the participants to set up scenes
become mo by demons trat·mg the simila rities and
question 0f generarlzabil. ity gs. For t h e '
tl� e bel'ng' how ever , most of the studles ' where 'circular questioning', 'live open supervision' and 'hypothetical future­
across a num . ber of se ttm . oriented questioning' were done? The study showed how these practices
on institutional mteracfIo n are more hke case stu d'les ed on data colle.cted fronm are made possible through the very details of the participants' action.
' are bas .
studIes usua Y
'
les o n 'nsti
1 tuti ona l mter
' ac t lOn As possibilities, the practices that I analysed are very likely to be general­
Case stud num b er 0 f b'Jects involved in such
. The essional
one or a few
is relatively �
sites
mall.
only
he p oble y �
! . : �:� s e�ific rofessional theory: fo �ple,
e articularly acute if the profexa
r
izable. There is no reason to think that they could not be made possible by
any competent member of (at least any Western) society. In this sense, this
I.S m�orm
� ork elf , solution-oriented the rapy (see
dIed study produced generalizable results. The results were not generalizable as
practice that
psychothe rapi
IS stu
sts work m � m th � f r a � e
distinctiv different
ely descriptions of what other counsellors or other professionals do with their
rIer:t� �' n ways that are ing) or
1991 ) ' n" 2003 ' Perakyla' forthcom
inte ract wIth th el r c clients; but they were generalizable as descriptions of what any counsellor or
e . g. Gale, other professional, with his or her clients, can do, given that he or she has the
l ana iy sts ( V e h VI l ame 1993) .
from tho se
therapists
of
with
psyc
ye :�
l �
If ! ere :
nt t ��� efl al inclinations (se e
a �
e.g . Bu
whether the results pres �d �111
ttn
ent
y,
d n such
3

same array of interactional competencies as the participants of the AIDS

For these rea sons , It IS I mpo rta that is saI case counselling sessions have.
an way ge nera ' ble Do es everything
rlza si te that
studies are in � . ractlOn '
. applY exclusively to the parti cula r
studies on i nstit utio nal mte ance?
was observed, or do th. .e re suIts have some WI'der relanevdin g of generalizablIt1 y,a
nal 'd' Ist rib uti onal' un de rst
ItIO one or
On cannot 0ffer m uch . Studying
In term s of th e trad .
ies on i nstit utio nal inte ractl
case stud
297

296
 RELIABILITY AND VALIDITY IN RESEARCH BASED ON NATURAllY OCCURRING SOCIAL INTERACTION

ANSSI PERAKYlA
VALIO/TY
Bearing these restrictions in mind, statistical analysis may be useful in
particular conversation analytical research designs, These include research
Quantification
nd � an fI fica tion inv olve s ano the r km
,
d of strategy designs that concern relations between distinct interactional variables in
dat ab ase s �
Use of large the generahzablhty (�nd also other aspects of the Asdity)
<:I vali of the standardized forms of encounters (like the medical peer review studies
for ensuring analyt, ca1 �esearch fmd'mg.s (ten Lla l ve, 1999: 144- Heritage by Boyd, 1998), or historical change in such encounters (like the changes of
dle, d by
8),
conversation contnb� utIOn to this book, some of the practicesFostu r exam ple,
presidential press conferences studied by Clayman and Heritage, 2002),
or relations between social categories and interactional practices (like the
shows in hisn ana1yst� 1end themseIves t 'codl'ng and counting', 0 , �esI-'
_

conversatio and Hentage, s (2002) recent study on question desi gn m p relations between gender and interruptions studied by West and Zimmerman,
in Claymaness conferences m, the Unl'ted States the journalthey ists' que stIOns 1985), (For a more thorough account on this, see Heritage, 1995), In any case,
dential pr regard'mg the degree f ,adversari�lness' that of eques
0
xhibited,
, however, the backbone of conversation analytical work, also regarding
were coded were rnad e tO sho w how t h e re 1at1've proportions
tIOns, generalization of research findings, involves qualitative case-by-case analysis,
CalculatiodIffe ns, . rsa na, 1ness, changed over , time, It was
re nt d egr ees 0 f a dve
showing t the journalists have beco�e much less deferentIalof suc and more
shown tha , t el' r treat ent of the presldent , Another example 2001)cessful dy
CONCLUSION

aggressive mon�IS offered�m Boyd's (1998'' see also Heritage et al., ns betstu ween At the beginning of this chapter, I pointed out that the specific techniques
quantificati , p eer reVIew ' , She stu .dIe' d te1eofphoanneinscon sultatio ny, Each of securing reliability and validity in different types of qualitative research
on medIcal me dica l repr ese n t a t '
IV es ura nce compa a pro- are not the same, The aim of this chapter has been to give an overview of the
th e 0
ncial coveragequafl'Itat1ve
hysicians andle, I ded a decISI
Pcons " On co ncer nmg' the fina ' imperatives faced and solutions found in conversation analytic research,
ultatio,n Y rat'lOn, Usi . ng qua n t't
1 a t '
lVe tech niq ues along with the f'1rst especially when such research focuses on institutional interaction, At a more
osed surgICal ope
Pones wed that thepint eractI' Ona1 format . of the init iation of
n co ncer ,
nmg general level, however, the considerations of validity and reliability in CA
, Boyd showas t 0
or f the out com
a strong rethd'eICop mng of the call set the trajectory for thee (the deci sio are similar to those in any other kind of qualitative research: all serious
topic of the call ords, ,
�1m, tla, tl'on, the participants were either qualitative research involves assuring the accuracy of recordings and testing
the surg , ery) ," In rawresu
othe
lt 0 f t h e t the truthfulness of analytic claims, In terms of the division of qualitative
as
ensumg revorIew'bu, reaucratlca , lly' ar19ned, and these alignments Ied to dI'fferen methods into three main 'branches' suggested by Silverman (2001), it seems
'collegially' ntitati�e that the specific constraints facing CA are closer to those of observational
decisions, o issues are crIt, ical , regarding the applicability of quaractic es IS research than those of text analysis, The questions about the quality and
At least tw, CA. F'Irst, straight£orward cod1'ng of interactional pthat large inclusiveness of recordings, for example, arise in both, and deviant case
techniques m OSSl'bl e, Many practIce , s mvO ' 1ve such(and complexity excIUSlve ', ) analysis is also used in both,
not al wa ys Pcases cannot be subsumed under S1' mpIe mutu ally Kirk and Miller (1986: 21, 42) point out that in conducting and assessing
numbers, of If comp1ex cases are forced under simple categories,consoSI'dmeth mg
categorIes , , ally lm from Sl' ght , This kin d of " era IOhtn t' qualitative research (particularly ethnography), the primary emphasis has
that is analytlcgloff (199 ' por t a nt m ay b e I os t that interactIOn mig usually been laid on validity rather than on reliability, whereas in quantitative
led Sche 117) to � rop o se the possibility only' and not o rderl y, in research the emphasis has been on the opposite, Put in simple terms, this
has 3 :
0 f th e smg u l ar occ ur renc e may imply that qualitative research is well developed in terms of validity
be orderly 'atwthe 1eve 1
agg rega : l i The other prob le m co ncerns samplinSIS,g and underdeveloped in terms of reliability, I hope to have shown in this
any relevant200ay,1: 249) at the er �r ��: q' uantitative analysis to provide a ba chapter that this is not the case with CA: CA can be considered a serious
rma n,
(Silve zation, the, sele In or� e t b e stu died should folls, ow adequate
ctIO 0
n f attempt to develop a method for the analysis of social action that is able to
for generaliproc�dures so as to e s�r: �h�ir re resentativenes Incolstulectld�esO ofIS combine concerns of validity with those of reliability,
statistical g lIke an o sam �in is rarely �ossible, The data ing medI�Ca A serious concern about the reliability of observations was at the very core
CA, anythinus and �1l1stltu � n:tIOnal�on�itions too strict, In research to workl of the initial motivation of Harvey Sacks in beginning the line of research that
too lab ns or psychotherapy, f ex� pIe' the researcher may havnebetween
orio we now call conversation analysis, The reliability of recordings remains
consultatio of data that he or s�e w��get access to, If the relatio if they are an inherent strength of CA -but as I pointed out earlier in this chapter, audio
with the kinand d opu lati on re mam s u ncle ar tests, (as
, statisticalstica 'In or video recording per se does not suffice as a guarantee of the reliability
le t h e p
the samp yield results that shoUld b e understood heuri llyqua only iv, e of the observations, The researcher needs to pay attention both to the technical
used, m 8, 2002), is does r:�t ne d to be a reason not to port nti
ay use tat quality and inclusiveness of recordings and to the interplay of spoken
Perakyla, 199at all, but !1 �IS a conSI era;1on that restricts their im in terms language with other modalities of communication and social action,
techniques ility of fm� dmg , s,
of generalizab 299

298
 II

RELIABIlITY AND VALIDITY I N RESEARCH BASED ON NATURALLY OCCURRING SOCIAL INTERACTION

VALIDITY
ANS SI PERAKYLA
. n 0f the researcher's analytic claim, s .in all Carmines, E.G. and Zeller, R.A. (1979) Reliability and Validity Assessment. Beverly Hills,
0
CA: Sage.
f vali
The m ain procedures�searc� Ill . :l� � the anal sis of the next spea1(er s III�tlOn
datlO
� .
er-
Clayman, S.E. and Heritage, J. (2002) 'Questioning presidents: Journalistic deference
conversation analytic�rdIllg actl. O devian?case analysis. InWcOl :versa and adversarialness in the press conferences of Eisenhower and Reagan', Journal of
the preC ne dlm ensl�ns
pretation of
� 0 r: tuti o nal int era ctio n,
�lms
de the validati?n of thetheclIssue
insti Communication, 52 (4): 749-75.
analytic studies whialso ch foc �
The se in clu Clayman, S.E. and Maynard, D.W. (1994) 'Ethnomethodology and conversation
ans e � .
of valida. tion have ce of an IllSt'tI t'10nal context of interactlOn, and analysis', in P. ten Have and G. Psathas (eds), Situated Order: Studies in the Social
. Organizaiion of Talk and Embodied Activities . Washington, DC: University Press of
concermn� the relevanthe results of case studIes.
U

of gener al Izab '1'ty

I I of America. pp. 1-30.
Dingwall, R., Eekelaar, J. and Murray, T. (1983) The Protection of Children : Oxford: Basil
Blackwell.
Drew, P. (2003) 'Comparative analysis of talk-in-interaction in different institutional
. e and Johanna Ruusuvuori for their settings: A sketch', in P.J. Glenn, C.D. LeBaron and J. Mandelbaum (eds), Studies in
. Davi d Silv e rma n, Joh n He ntag Language and Social Interaction: In honor ofRobert Hopper. Mahwah, NJ: Erlbaum.
I wIsh tO thank of this chapter. Drew, P. and Heritage, J. (1992) 'Introduction: Analyzing talk at work', in P. Drew
comments on theCearlier versions
*

ana IYSIS, see ten Ha

ve (1999), and J. Heritage (eds), Talk at Work: interaction in Institutional Settings. Cambridge:
. ns n th e th od of conv ersa tion ehr (1 997) , Cambridge University Press. pp. 3-65.
1 . For general diSCUSSlO l°- utc�:y and Wooffitt (1998), Pomerantz and F
"

) Drew, P. and Sorjonen, M.-L. (1997) 'Institutional dialogue', in T.A. van Dijk (ed.),
Heritage (1988, 1 99and WIootton (1989) .
5 ,
Discourse: A Multidisciplinary Introduction. London: Sage. pp. 92-1 18.
Silverman (1998)m. orga uza . ' c , see also Boden (1994). Fairclough, N. (1999) Critical Discourse Analysis. London: Longman.
2 . For temporality ' �On tonlOn
Sl t
s

h e s tu d y of i nte ractions that are informed bY Firth, A. (1995) "'Accounts" in negotiation discourse: A Single case analysis', Journal
3. For a generaI d I
CuS and Vehvilainen (2003) . of Pragmatics, 23: ] 99-226.
professional theon� es, see P era"kyla Gale,J.E. (1991) Conversation Analysis of Therapeutic Discourse: The Pursuit ofa Therapeutic
Agenda (Vol. XLI in the series 'Advances in Discourse Processes'). Norwood, NJ:
Ablex.
Gergen, G. (1994) Realities and Relationships: Soundings in Social Construction, Cambridge,
MA: Harvard University Press.
Glaser, B.G. and Strauss, A.L. (1968) Time for Dying. Chicago: Aldine.
Goffman, E. (1983) 'The interaction order', American Sociological Review, 48 0): 1-17.
Goodwin, C. (1992) 'Recording human interaction in natural settings', Pragmatics, 2:
181-209.
Goodwin, C. (1994) 'Professional vision', American Anthropologist, 96 (3): 606-33.
Goodwin, C. (1995) 'Seeing in depth', Social Studies of Science, 25: 237-74.
Goodwin, C. and Goodwin, M.H. (1997) 'Formulating planes: Seeing as a situated
activity', in D. Middleton and Y. Engestrom (eds), Cognition and Communication at
Work: Distributed Cognition in the Workplace. Cambridge: Cambridge University
R EFE R E N CE
S
,
. Press.
and Cultural Studzes
Hak, T. (1994) 'The interactional form of professional dominance', Sociology of Health
Alasuutan. , P . e. a rchin g Cu lture : Qualitative Method
. n (1tatiivisct
(1995) Rese
London: Sag me
and Illness, 16 (4): 469-88.
t ' n, S . and YI"ostaI P (1994) Hammersley, M. (1992) What's Wrong with Ethnography: Methodological Explorations.
,
Sosiaali tutkImuksen Kva
Alkula, T., P"onPorv London: Routledge.
.

oo: WSOY. (199�) ,Criteria for assessing interpretive vahd.Ity l

0,
. . ·n

Men telmat.
and JohnsonN, J.KM. De . Hammersley, M. and Atkinson, P. (1983) Ethnography: Principles in Practice. London:
Altheide, D.L. d Y . S. Lincoln (eds), Handbook of Quallta Tavistock.
t1ve
.
rese arc 1 " , m . nzm an
ualitatlve ten Have, P. (1999) Doing Conversation Analysis: A Practical Guide. London: Sage.
usand O�ks, CA:J, sears)' �f9 84)
1
.
l,esearch. Tho
485-99
� Heath, C. (1992) 'The delivery and reception of diagnosis in the general practice
'

StudIes
. e U11lV. ersit' y P ress.
m
es of Social Action :
Atkmson, J .M . Anaand H entage,bnd, Cge. Cambridg Str ctur
Cam ,' in Action. Cambrid ge:
Poh.ty. . consultation', in P. Drew and J. Heritage (eds), Talk at Work: Interaction in Institutional
Settings. Cambridge: Cambridge University Press. pp. 235-67.
quaIs'" . initiating dlSCU sSlOn
Conversation lysi s.
Bo den, D . (1994)'BuThe e
ons
f
izati
pany
n
. Heath, C. and Luff, P. (2000) Technology in action. Cambridge: Cambridge University
Orga
' "com
.
' the
of Talk
t
Busmess
Boyd, E. (1 998) icalrea au h
ucrah� : ��r��an Sociological Review, 63: 200-24.0

Press.
du rin g med peer reVIew . ication. Newbury P
ark, CA: Heritage, J. (1988) 'Explanations as accounts: A conversation analytic perspective', in
Y mun
Buttny, R . (1993) Social Accoun
t a b1T I t in Com

Sa ge. 301

300
��------

ANSSI PERAKYLA VALIDITY RELIA BILITY AN D VAL I DITY

IN RES EARC H BAS ED
"

ON NATU RALL Y OCC UR "

RING SO""IAL INTE RACT I O
N
C. Antaki (ed.), Analysing Everyday Explanation: A Case Book of Methods . London: Perakylii, A. (forthcoming) 'Ma king
Sage. pp. 127-44. Conversation Analytical view' Psychlinlt�S 111. psychoanalytic interpretations: A
Heritage, J. (1995) 'Conversation analysis: Methodological aspects', in U.M. Quatshoff Periikylii, A. . and. S·l1 verman, D.' (9 ( ) erapy Research.
(ed.), Aspects of Oral Communication. Berlin: Walter de Gruyter. pp. 391-418. C.. ommunication formats in AIDS cou91 ) 'Reth·, 1111<l·�g sp eech-exchange systems:
Heritage, J. and Atkinson, J.M. (1984) 'Introduction', in J.M. Atkinson and J. Heritage Peraky la, A. and Vehvilainen, S. (2003) ,n��II"111g , S?clOlogy, 2� (4): 627-51 .
(eds), Structures of Social Action: Studies in Conversation Analysis. Cambridge: stocks of interact nversatlOn �nalYSls and the professi
Pomeran tz: A . (198io0)nal'Tekno wledge': D ISCou onal
.
Cambridge University Press. pp. 1-13. II" 111 g my SI de: "Li
rse and Society, 14 (6):
727. -50 .
Heritage, J. and Lindstrom, A. (1998) 'Motherhood, medicine and morality: Scenes from mit e access" as a "fIshing device''',
Pomerantz, A. (1990y,) '50 : 186-98.
SoCl·ologlcal lnquir d
a series of medical encounters'. Research on Language and Social Interaction, 31: 1 Ity and general"·l�abTt
295-310. methods : ConversatOn ion
the va l·d·
analy tic cla im 1 1 Y 0f conversation analytic
Heritage, J. and Watson, R. (1979) 'Formulations as conversational objects', in G. Psathas Pomerantz, ' s'
.

A.
. actioand Fehr B .J . (199 'ConVersatlO anaIYSls , Co m umcatlOn Mono
graphs, 57 (3) .· 231 -5 .
n as sens' e making7)pra .
�
(ed.), Everyday Language. Studies in Ethnomethodology. New York: Irvington. of SOCIal · : �n app
. (ed.),roDIsco
�ch to the study
ctlces�' in ;.A. van Dl)k
,

pp. 123-62. InteractlOn. London: Sa

Heritage, J., Boyd, E. and Kleinman, L. (2001) 'Subverting criteria: The role of precedent Sacks, H. (1984) 'Notes onge.met pp. 64-91 .
hodology', in J !"1 Atl(lp. son and J.
urse as Social

in decisions to finance surgery', Sociology of Health and Illness, 23 (5): 701-28. Structures of Social Acti
. on: Studies in conversa t · Heritage (eds)
Houtkoop-Steenstra, H. (1991) 'Opening sequences in Dutch telephone conversations', Umver · slty I)ress. pp. 21-7. IOn Analysl·s. Cam b ·d
ge: Cambridge'
·

in D. Boden and D.H. Zimmerman (eds), Talk and Social Structure: Studies in Sacks, H.,. Sch eglo ff,
n

Ethnomethodology and Conversation Analysis. Cambridge: Polity. pp. 232-50. orgamza tlOn E . A .
. of turn- and Jefferson' G (97 4) 'A SIm . plest sys
takin g for conv ersa t:l�n , tema tics for the
Hutchby, 1. and Wooffitt, R. (1998) Conversation Analysis: Principles, Practices and Schegloff, E.A. (968) 'Sequencing. 111. conve , Lan guage, 50: 696-735 .
Applications. Cambridge: Polity. 70: 1075-95. rsatIOnal openings', American Anthropol
Kenny, A. (1973) Wittgenstein. London: Allen Lane. Schegloff, E.A . (1986) 'The rout·me as achle . . vement', I:l.
ogist
'
Kirk, J. and Miller, M.L. (1986) Reliability and Validity in Qualitative Research. London: Schegloff, E.A. (987) 'B etween macro an uman Studies, 9: 111 -51
Sage. J . Alexander, B. Giesen, R. Munch and d micro·· Contexts and other conn ectio.
Lindstrom, A. (1994) 'Identification and recognition in Swedish telephone conversation Berkeley and Los Angeles: University N · rS;nel�er (eds), The MicrO-Macrons', in
openings', Language in Society, 23 (2): 231-52. Schegloff, E.A. (991 ) 'Reflect·IOns on talk Ca 1 ornJa Press. pp. of 207-34. Link .
Luff, P. and Heath, C. (2002). Broadcast talk: Initiating calls through a computer­ Z·Immerman (eds)' Ta/f" and S OCIa. I St and sod , .
aI structure , m D . Boden and D. H .
mediated technology. Research on Language and Social Interaction, 35 (3): 337-66. Conversation Analysis.
Cambridge . P n ructure·. St u dles · m. Ethnomethodology and
Maynard, D.W. (1991a) 'Interaction and asymmetry in clinical discourse', American Schegloff, E.A . 0 992a) 'Intr oductio�1'' �� �.Pt���s, Lectures on Con4-70 .
Journal of Sociology, 97 (2): 448-95. ed. G. Jefferson. Oxford: Blac
Maynard, D.W. (1991b) 'The perspective-display series and the delivery and receipt Schegloff, E.A. (1992b) 'On talk and kwell· .X Ixl1. versation, Vol. 1,

of diagnostic news', in D. Boden and D.H. Zimmerman (eds), Talk and Social (ed s), 7� lk at Work: Interaction itsin iIts� nstit���;�l �e�:mgs. Ca
.
�� :- o c lOn , in P. Dr�w and J. Herit
,
age
Structure: Studies in Ethnomethodology and Conversation Analysis. Cambridge: Polity . Umv. erslt y Press. pp. 101 -34. mbn dge: Cambridg e
pp. l64-92. Schegloff, E.A. (1993) 'Reflections on .
Maynard, D.W. (1992) 'On clinicians co-implicating recipients' perspective in the uage and Social lntera�
u antif lcati 0 1n the study of conversation',
tion, 26 . 99-�2 8 .
Research on
delivery of diagnostic news', in P. Drew andJ. Heritage (eds), Talk at Work: Interaction SeaIe, C. (1999)Lang The a lI"tYtf quallta . .
tzve research . London :
in Institutional Settings. Cambridge: Cambridge University Press. pp. 331-58. Silverman, D. 0985rQuaII atlve Metho Sage.
Maynard, D. (1996) 'From paradigm to prototype and back again: Interactive aspects S·lverman, D. (997) Discourses of Cou dology and SOClO. Iogy. Aldershot: Gower.
of "cognitive processing" in standardized survey interviews', in N. Schwarz and Silverman, D. (998) Harvey Sacks and nsenmg. . London: Sage .
1

S. Sudman (eds), Answering Questions: Methodology for Determining Cognitive and S·l1 verman� D. (200 Conversation A na IYSI.S.
1) Interpreting Qualitative Data.. Methods Cambridge : Polity .
Communicating Processes in Survey Research. San Francisco: Jossey Bass. �nteractlOn, Second Edition. Lond .
SmIth, D. E . (1974) 'The social constr on. Sage.
for Analysmg Talk, Text
Miller, G. and Silverman, D. (1995) 'Troubles talk and counselling discourse: A uc t" IOn 0
and
comparative study', Sociological Quarterly, 36 (4): 725-47. �4 (4): 257-68. f doc ume ntary reality', SociologicaLInquiry,
Parker,1. (1992) Discourse Dynamics: Critical Analysis for Social and Individual Psychology. SmIth, D.E. (99 0) The Conceptual P ractl.ces
Press. of power. Toronto:
.

London: Routledge. University of Toronto

Perakyla, A. (1995) AIDS Counselling: Institutional Interaction and Clinical Practice. Strauss, �.L., Fagerhaugh, S., Sucz
Cambridge: Cambridge University Press. of MedIcal Work. Chicago: University
eck, B. and Wi . go ener, C. (985) The Social Organ ization
Perakyla, A. (1998) 'Authority and accountability: The delivery of diagnosis in primary Sudnow, D
. (967) Passing On: The Social Orgamz Press. .of Ch Ica
health care', Social Psychology Quarterly, 61 (4): 301-20. Prentice Hal l. atlOn of Dymg. EngleW
OOd Cliffs, NJ .
Perakyla, A. (2002) 'Agency and authority: Extended responses to diagnostic statements Veh. VIrame · n, S.
(2003 ) 'Pr epar in g .
Interaction', Text, 23 (4): 573-606an. d delivering mterpretations in psychoanalytic
.

in primary care encounters', Research on Language and Social Interaction, 35 (2): 219--47.
302

303
 II

ANS SI PERAKY L A 1) VALIDITY

. , .
lan gua ge and dIscourse , m T . A . van
' Vo 1 . IV . London: Acade
nd er
ourse Ana lYS1S,
D.H . (198 5) 'Ge ' mic Press.
West, C. and Zimm erman,
d . ) , Hand book of DIsc
1 (e
D"IJ1< . .
pp. 103-24. r P ro du ctio n: Co mp ute r-a ided dIs patch m
. 5) 'A tech.nolog y of orde
Whalen, J . (199 . ns' , 1· n P. ten H ave and G · Psathas (eds). ' Situ. ated Order:.
IO
mu niCat DC.
public safe ty com .
nizatIOn 0f Ta IIc an d Embod ied Activities. Washmgton,
Studies in the Social Orga
University Press 0f America umous w n't'mgs bY
187-230.
. pp
Wittgenstein, L. (.l Blackwe 11 . . , .
.
Remar,cs. Ed1· ted from his posth
. 1 '

975) Ph Iwsophlcal "

1 versation anaI yS1S , In

�

eaves and R . Wh ite. Ox for d:

on the . m � h d 1 Y of con
trans at ed by R · Ha rgr
�: � ��
R. Rhees and
A.J. (19 89) 'Re marks C evedon:
W ootton,
ConversatIOn. In e iscip i ary Perspective. l
l n
D. Roger and P. ' Bull (eds) , '

38-58.
Multilingual Matters. pp. 2

Michael Bloor

This chapter explores two case studies which provide illustrative details of
two different but related approaches for researchers who wish to address social
problems and who are also sceptical of the possibilities of extensive influence
among the policy-making community. Both of the approaches aim to influ­
ence practitioners rather than policy-makers and both link particularly well
with qualitative research methods. In the first case study, an ethnographic
research project is viewed as an analogue or partial paradigm of successful
practitioner work, in this case outreach work among male prostitutes: in effect,
the ethnography may be viewed as a demonstration or pilot outreach project.
In the second case study, ethnographic work provides the material and the
stimulation for practitioners to evaluate and revise particular facets of their
own service provision.

SOCIAL RESEARC H E RS AS SOCIAL E N G I NE E R S?

It was rather a shock for me to read in Carey's (1975) social history of the
'Chicago School' of sociology that in the 1920s the foremost practitioners of
the foremost school of sociology were divided about how sociological
knowledge should be applied. Should it be used to influence policy-makers?
Or (and here lay the surprise) should sociologists intervene in social prob­
lems directly as consulting professionals, like clinicians or architects? I was
vaguely aware that some hundred years earlier Auguste Comte had proposed
a similar priestly cadre of sociologists to direct society along enlightened
(and Enlightenment> paths. But the realization was somehow monstrous that,
as late as the 1920s and contemporaneously with, say, Eliot's 'The Waste Land'
and a hundred dystopian diatribes, my intellectual forebears could hanker

304
 ADDRESSING SOCIAL PROBLEMS THROUGH QUALITATIVE RESEARCH
ITY
MICHAEL B LOO R
THE WIDER COMMUN
ir nostrums : with the French 'Encyclopaedists' of the eighteenth century, namely that social
en g . r social life and institutions to thewar
mee ed t�em. life can be improved by planned intervention derived from accumulated
re- th
after the power to that pow er, rather than humility, which stitutI: Onalized
of in scientific knowledge, itself the product of social research. In late modern
It was the absence ( 975: 71 94) , It. w as th . lack of the kind of rather than a�y society all three tenets of the Enlightenment paradigm have come under
in Carey's analysism1edi cine exercIses
-
. 0 :r a lay clientele, 'postmodernist' criticism: it is no longer universally accepted that planned
etence: whIch
a utho rity whi ch .
m k o wle � g: or in technical comp ts �nd intervention is capable of bringing about desirable social change, or that
acknowle dged defi cien cy
hIC
�
h soc lOl0 gy wo uld develop. iologl�entists
Soc
scientific knowledge can facilitate this, or that social research can produce such
.
determmed the p ath alo ng W stalls as sCI. .
� a 1:e
1 pted to set out theirpar � knowledge. Hammersley'S review of these postmodernist criticisms leads
ea
other social res sIOrc� e rs ev ent
t � est was largely s ed . the dlIec lve him to the assessment that they serve to qualify severely, rather than demolish,
rather than pro fes ialnals
, an
( t h e peo p les of the Soviet Umon were ess the possibility of a social impact for social research: the scope for and feasibility
. tervent10' n of soc exper ts
m of successful policy intervention has been overestimated in The Enlightenment
fortun ate). Project and the role of research in bringing change about has been exaggerated
and misunderstood, but this does not mean that soc'ial improvement is
impossible or that knowledge lacks all authority.
SO C I AL
E RS?
P OLI CY- M AK .
.lImIte on of soc iologi sts to inf luenc� ?ohcy-
irati . ted out by various cntlcs that.
· the· 1960s' the more ' tdaCasp A PRACTITION ER A U D I E N C E FOR SOCIAL R ESEARCH?
Smce d k It was pom
makers has also be�Ity n un er at
so ur h � P olicies fro m res earchers: instead, researctoh The policy community is not the sole audience for qualitative social research.
the policy commun one rarely
co n lI m � prefe rre d policy option, or perhaps �e Uses On the lay audience for qualitative social research, see Silverman (2000: 275-6).
d to tion. Bulmer (1 982) , ine Ttlal role
ould be comm issi inC .
:elay a necessary but OnV enIen �� � ��:ht to redefine an infl� � ( 9 )
e Sociologists who have conducted research on sociological aspects of health
: � �� : Taking up J anowItz � po
as on of t e 1 :2 and medicine have long been aware that there is a role for sociologists as
of Social Resea rch, � ese criti ci ms. 1ICY participants in debates on public policy, but that there are also audiences
a � .
ce ightenm ent' modelspo0hcy
for social science mn the , eng m eer mg ' a nd 'enl . , but of practitioners (clinicians, nurses and other health professionals) for social
distinction b etw ee the res e rch cannot engineer cha nges of
res earch, Bul m er arg ue d t � at �
. ct Imp act on the polic y clim ate through pro­ research. Practitioner-oriented social research has also been the subject of
ortant . m dITe
provl' dl' ng descriptivethe accounts revisionist criticism. One strand of such criticism is that the researcher becomes
It. can have an imp '
lO n an d infl ue nce, , state the ally of the practitioner in exploitative relationships with patients or
cesses of intellectual rpr asso � cIa
etatlO�s. �lve
. �: ; n (200 1) has ter me d this
on 'the uses t clients, servicing the practitioner in practitioner-client relationships which
and theoretical inte has gent � n� �. C� how Bulm er'S bookial earch or reinforce patriarchy, say, or white supremacy. The other strand of criticism
counsellor' role andns out to e so e y about the uses of soc res of practitioner-oriented social research is that articulated by commentators
social res earch' tur b een under such as Schon (1983), who have followed Schutz (1962) in arguing that profes­
p olicy-makers. tenment' al d ,e . rin ' m odels have long
Both the 'enligh ; ltl
C � . '�����:ial�esearch' model. B � cker (1 967),
arg u;d
sional work does not entail the deployment of scientific knowledge, but
rather involves the deployment of a different kind of knowledge altogether,
of the an d
attack from advocatese rheton. ca que tion 'whose side are we on? research, . or
for example , posed one th. : y-oriented knowledge-in-action, which is rigorous but not comprehensive, task-oriented
the case for act ion - nte d r ; � ea: ather than policpatio
� emanci n rather than PO�lC� but not systematic, and experiential rather than research-based. In this reading,
.u�h:r
ough
progressive social chan ge
for
ac
a
11
pa
;�san sociology that spoke up for the
thr
rs m
social research has little of value to contribute to practitioners' work. However,
such criticisms hardly apply to that social research which takes practitioners'
initiatives. He callelited
elIt . es wh ' ch w ould include policy-m ake answer, everyday work as its topical focus: social research which seeks to describe
do s ag ain st the es , . On �IS . del b elieved to defy a simple res �
ear
nu�ber . Today Becke
estl and compare practitioners' everyday work practices self-evidently invites
. enc . ng cross��tti:g responsibilities to eagues l. t �
r ' s qu thei r
pen � to t�_kee ers and to their coll � practitioners to juxtapose and weigh their own practices with those reported
with researchers ex agenCIe S,
� � e mt� r
�
by the researcher. Qualitative research techniques, with their capacity for
subjects, to fun din ?
ty. A nd Bec\ :;, s rh et
scientific comn:unI em r��. 1 g � t Silverm an (200 1 : 260) has nd s on)
ric of sid es is tho ug ht
call
to
ed Its pr rich description, are favoured techniques for research focused on everyday
� �
lectually disablIng, ea e rut ' he plight of the underdog, a (eng�mee;ing,
work practices.
The first case study reported in this chapter is a street ethnography of HIV­
commitment to a rev models
995 has ch racterized all thsree ightenmen: �odels tin related risk behaviour among Glasgow male prostitutes. Safer and unsafe
Ham mer sley (1
cnt
1 .
Ical ) as � iffe
nligh tenment and they all endorse certaln Enlightenment id
ren t vari
.
etie of Enl
eas o ngm a g commercial sexual encounters were compared: unsafe encounters were found
:ith a capital E, sin ce
307

306
 II

ADDRESSING SOCIAL PROBLEMS THROUGH QUALITATIVE RESEARCH

MICH AEL BLO O R w THE WIDER COMMUNITY

ter. by the clients of pros-

The harm reduction approach
. ntr ol of the se xu al en c oun rcised
to b e assocIated WI'th co . d ith particular techmques 0f Power exe The HIV epidemic, along with its toll on lives and health, represented an
' e lion
. opportunity to change the policy climate in respect of male prostitution. The
� �s�� ��e:��::S:�����:;:
tut�
: in ! :
:l P ��� ��i� ����;:� �:� ::�Od�
icate pOSS li
! situation was analogous to that in the drugs services, where a range of services
f;r�hose engaged in e l ;s��: ��r :::�eSig� of successful outrea�� :a�;a-
sex ua (most notably syringe exchanges) was put in place for existing drug injectors

III thl' c'

o !
� nd e���:�n��e ��o:d �ase study i.s a com��::t�: ��n;e�:�; hlighted
p t . �g who were not motivated to abstain from drugs or to change their route of
administration from injection to smoking or ingestion. This new drugs policy,
. ity practIce . Th e
tions i n th eraP eu
�' ��� �� �
� d the ra peutic practice foun�:� r:�t';�� ' lar com-
ea. rcher's
which became known as that of 'harm reduction' or 'harm minimization',
a number 0f feat d ted els ewh ere, w l . e argued that 'the spread of HlV is a greater danger to individual and public
munities that could ? S�IpS � � ��O �� therapeutic community prac ��
rof bl b health than drug misuse . . . [and that] services that aim to minimize H lV
fie ldw ork rel atl�� i ticular new m e�� :; s risk behaviour by all available means, should take precedence in develop­
close rs to exp eriment with par
encouraged the practItIOne ment plans' (Advisory Council on the Misuse of Drpgs, 1 988). In similar
working. fashion, it became possible to argue the case for services targeted at male
prostitutes which had as their priority not the elimination of prostitution, but
the minimization of individual and public health risks.

methods and service UnJVI�SIUn

These changes in the policy climate are clearer in retrospect than they
d erVlces for rna1e prostitut es hardly existed
for
The .
art�cula on� .�1��ld ��0t:
Prior to the HIV epidemic��r;;::bl were in the late 1980s when the fieldwork on male prostitutes' HIV-related
. th e UK . There were a f
: exce pti ons
. .
: in p r
1tr risk behaviour was begun. The study was part of a wider programme of
twise, and 1ts pIoneenng drop 1 research on social aspects of HIV / AIDS supported by the Medical Research
m
ase d ch arity 'Str ee i n Gla sgow.
the Lon��� � _

tai nly �here we re no ta ��;:ed ser vic es Council. The findings and the methodology of the study have been fully
male t tes. But cer an t en q about male prostitution
lndee�, when I and col lea gu es £lr st b e �
. l?Ism s me quarters about whether described elsewhere (Bloor et al., 1 991, 1992, 1 993). After pilot work, six
er ed sce ptiC � ought to different sites - two parks, two pubs and two public lavatories - were selected
in Glas ��; :� enctuount l� ��;Yow (it was thdle-
m

m r� u n ac ally e xis ted on an y sca

h
�: o n effete, mid class
for time-sampling; non-streetworking prostitutes (escorts, masseurs and call
b :�Kch more the sorttitutIO � �ighlY diversified activity: tl:e 'c�tlhl
of hin
� � th at m l men) were contacted through their advertisements and the study's own
l
' s
.
Edin�� rg�) ' M 1er:S�:fullY decorated flat, packing fo t ' to BraZlcal
n, IIIf ac t , l WI advertisement in the gay press. The ethnographic fieldwork was conducted
m an hIS 0:n , IS. f r re o ed from the group 0; s����ing , cat-. , ling, in pairs for security purposes. Prostitutes were contacted by a combination
� � �selves with their b ottles of Buc�le �'neof , of cold-contacting and snowballing: some of those contacted had never
m

a businessman client ml g previously spoken to anyone about their prostitution activities. Both gay and
jostling teenagers .war �I 1 � ot all m ale pro stitutes have muc n��ems self-identified 'straight' prostitutes were contacted, as were both drug injectors
outs�de the late-:���a�: :�ltiple and co:nplex proble: ��:;l !or;etime�
C �i e and non-injectors, and both novices and experienced prostitutes .
:,�:� � ;;��l:�;' h?USa���:��� C����::�� � �:':�:�;:� Ovide�s
d; ��� �
g
;��
While the fieldworkers' primary objective was research, it was recognized
from the outset that the fieldwork also offered opportunities for health pro­
unpre en�:������� has made special�st servi e :e:���� � f
� I � : w the th vaagteeen motion: relations between fieldworkers and research subjects can never
be scientifically neutral (Hammersley and Atkinson, 1 995) and an attempt to
�l���J mforaleh0pr1: ost itut es co nta
s a � ed
�
°
o�� � e�e rcts did not occur in pri � preserve a fictional neutrality should never be used as an excuse for failing
of consent �ose \:� a ;a;: J qualified toler�tio� to ;e ��::��:
xua
on was P �:�e� :: utionn.
OhC e �d °!e to attempt to save lives. The Greater Glasgow Health Board provided condoms
a1tho� .gh the P ost ituti
prostItutIOn, w l1Cre y male s tre et pr suitable for oral and for anal sex for the researchers to distribute (when the
e xte nde d to male r e p rostit m ale
suppressed, that toleov1��ers WI�t�:� �nterest in providing s�rvices for
ra n wa e fieldwork started, condoms suitable for anal sex were not freely commercially
available); an advice leaflet was also handed out which gave advice on HIV
And m ale service prnerable. to mlSC . onstructions of their motIVes.
1
wer e vu prevention and also gave contact numbers for HIV / AIDS counselling and
prosti tute s for other relevant services such as welfare rights and homelessness.

309

308
 Rli.

ADDRESSING SOCIAL PROBLEMS THROUGH QUALITATIVE RESEARCH

MICHAEL BLO O R THE WIDER COMMUNITY

. (Davies and Feldman'
¥J

encounters with regular clients than with casual clIents

1991), possibly because sexual e ncoun ters with street . wor kers' regular clients
S tudy fin din gs
ndo m s and adv ice leaf lets could be thought to generate are more .11'1(e1y to take place at the clients' h ouses WIth more attendant privacy
If the handing out of ouraging the reporting of unsafe com mercial sex, then
co
d of than semI-public locations such as �� parks and back lanes, where disrobing
a 'reporting bias', disccan only have been mar ginal because at least a thirtheir may be difficult,. uncomfortable an angerous ' prostitutes and clients for
men t
such discourage ntacted reported uns afe sex with at least some rence of The subst't1 utIOn of overt negotiations between
tho se prostitutes co tners (unsafe sex was defined, following the Ter ter the furtive and largely non-verba1 exchanges charact.' ens. t'IC 0f- many encounters
current commercial par with out a condom, because of the grea would have advantages beyond th prev ' �ntlOn of ElV infection and of other
sexually transmitted diseases On: c?�sIderable advantage might lie in an
ana l sex with or ed with
Higgins Trust, as nsaf e com mercial sex was associat ercial
attendant reduction in the lev�ls of �����ce surrou�ding male prostitution.
re in a nal sex). U m m
risk of condom failu st to female street prostitution, where safer co of
client control. In contrapractised and the women assume directive control R�pes, muggings and assaults (of l s by prostItutes, of prostitutes by
sex is almost al way s
and Barnard, 1992) , in many
male prostitu te- client clIents and of both prostitutes a,�d crlents by 'queer bashing' third parties)
the encounter (McKeganey who assu mes con trol and deci des on m atters such are commonplace ' during the:IX ee�-month fieldwork period three of our
encounters it is the clie nt
loc atio n. Safe r com
utes
mercial sex among m ale provestitaway thirty-two resear�h subjects ere c lar?"ed with assault and a fourth was
and its
as the type of sex particular strategies of power to w rest the ini tiati imprisoned. Many (but not all) of th ese VIOlent altercaf '
IOns are d Isputes about
was associated with payment up front (univers ally practised by female m oney. There were no 'goin g ��tes' for the vanous . sexual services on
. agreement offer.
from clients. Seeking such successful strategy. Getting payment up front prostitutes took what money ley cou1d and' without pnor on
prostitutes) was one the clients, who feared (with some justification) thats cha rges (sometimes without even �nor ' agreem�nt that a charge was to be
was not popular withdo a runner', but that the minority of m ale prostitute levied), the scope for violent dispu es was conSIderable, as is illustrated in
the prostitute might ' insist on prior payment were all currently practising the following field note:
in the sample who did.
safer commercial sex m oney up front was not the only successful counter-
.
['Sammy' said he'd] never been cheated aut of hIS money: he'd make sure he always
However, getting the er used by male prostitutes to insist on safer sex. got hIS money (this was said with a dd h�rd ��1 �hasis . . . ). He and 'Kenny'
:� ::
.
vailing strategy of pown a highly covert and ambiguouS activity, few words laughingly recalled an altercation wi on f Calm s punters [i.e. clients]. Colin
Male prostitution is oftenot even always clear to both parties that the encounter '
was demandi�g twenty-five pounds and the punter swore he was only due fifteen
.
are exchanged and it isSafer sex is likely to be associated with any techniqu es pounds, refusmg to hand over the extra ten pounds Kenny, m h'IS cynical way
Iined to believe Colin b t Sammy sa rd �e'd rather believe a mate tha�
.
one . and mak e "
l
1�
nter was dIsmc
is a comm ercia e amb igu ity tha t surrounds the encou following field some dirty old punter. Sammy hadu. tervened, whIpped a knife out and held it in
that serve to disp el th
atter for overt discussion, as in
the :
type of sex (and prices) a m �
front of his face (this was mimed u for o r benefit). The punter instantly pulled
out the extra cash, shot off and ha never een seen at the toilets since.
note : de
would come and stand besi
was to stand at the urinal . The client a nd he would Implications for service provision
His pro ced ure
the client would put out a hand nt. With the
was clear, . It. indicated how
es got
him. When the coa st would leave at that poi
sorry but I charge.' Some This research project had two possible p olicy pay-offs. FIrst,
immediately say 'I'm
a rate. H e would acce pt 1 0 pounds but sometim anal both unsafe commercial sex an d VlO ' 1 ence could be red uced, namely through
he'd neg otiat e nts ask ed him for
remaind er
ays did hand jobs or oral
sex . . . . If clie encouragmg ' male prostitutes to en�!� I'::;7rt negotiations with clients. And,
20 pounds . . . . He alw
sex he told them to eff
off. secondly, it indicated a possible i �
o� ha.t encouragement, namely
prostitute and client is onl y one o�treach work associated with cond om d'Istn utIon at regular prostitution
ove rt neg otia tion s bet ween n of safer SItes .
E ncouragement of ibly the most effective strategy) for the prom otio as to Outreach work, taking services to elients rather :han waiting for clients to
strategy (albeit poss ed w
by some respondents a 'book'
r str ateg y pra ctis attend at agencies, has a Ion historY �ut has been lIttle evaluated (Rhodes et
a1., 1991). It is clear that outr!ach work IS the only means of del'Ivering services
commercial sex. An othe sex by building up asual
attempt to screen out thwho ose clients looking for analise
pract d) and refusing allcoc nfine to cl'lents wh0 are unable or unW 'n' g o attend agencies and the HIV / AIDS
of regular clients (wit
cts.
h
But it
m
is
safer
diffi
sex
cult
was
for str eetworking prostitutes todent had �
epidemic has greatly stimulate� :� eveloP.n:ent of outreach services to
comm ercial conta 'regulars' (only one streetowork
ing respon
stigmatized and victimized 0 ulati�ns (drug I�Ject�rs, gay men, female and
them selves exclusively to ondents h ad some)
succeeded in w orki ng thu s, alth ou gh ano ther tw resp
me with additional casual contacts :
male p�os.titutes) who are j!d ed to b� at pOSSIble fIsk of ElV infection and
men ted thei r inco on in transmisslOn. At the time this study was conducted, Glasgow had no outreach
'regulars' and su
dy
pple
fou n d tha t anal pe netr ation was actually more mm co
and one UK stu 31 1

310
MICHAEL BLOOR 4 THE WIDER COMMUNITY
ADDR ESS I N G S OCIAL
PROB LEMS THRO U G H
QUA LITATIVE RES EAR
CH
project targeted at male prostitutes: there was a drop-in centre for female
prostitutes, but no outreach workers were attached to it and no men were
admitted to the drop-in premises. Ethnographic fieldwork, in its protracted
and regular contacts with research subjects, has much in common with services
outreach work and it was therefore possible for the ethnographic study to take
on the character of a local feasibility study for a male prostitute outreach Therapeutic comm unities are
found in a var'lety .
service, demonstrating to the sceptical that appreciable numbers of male and non-residential, long-.t , erm and s-hort-term)ofcashap es and sizes (residential
prostitutes were working in Glasgow, that levels of HIV-related risk behaviour groups (psychia tri c patients
ex-psyCh'l �tn. c p atIe ', term ' g f'or a range of client
were high and that outreach contact could be established, Moreover, the nature difficulties, adults with le;rni . dr nts, chil dren with learning
of the fieldwork contact that was established augured well for a future pns, oners, and so on) ng d1' fflCUl tIes ug u sers, and. alcoholics
, common an. , with a range 0f· d'Ifferen' t staffmg arrangemen
outreach service: large quantities of condoms were distributed (to clients as haVIn ' g In appr to theraPeu hc, work as an esse ts, but'
well as prostitutes); even highly socially isolated individuals with no contact activity which can transfoorach ntially co
with other prostitutes proved contactable; working relationships were estab­ Iavatory cleaning or compI e m" any m unda. ne event In ' the communitygniti ve
t
.m the lIg . ht amm g a bou the n OIse " mg that (be it
lished with important local individuals such as bar-owners and managers, of som' e therapeutic parad' gm .
' )'by redefm eve nt
toilet attendants and (at an appropriate distance) the police, and the project the paradigm may vary � (B loor e t al., 1988) , The nature of
proceeded with no threat to the safety of the ethnographers . of the event (as shOWingfrom com � � ty to com munity,
responSIb1l1�ty u s but the redefin n
Throughout the fieldwork period I had briefed public health personnel, pathogel1lc' way of rela ting ' ay, or s�ek111g' out a new anditio les
social work staff and AIDS charity workers about project developments and sets it apart and transforms to oth ers) as an OccaSIOn or a topic for therapys
provisional findings. At the conclusion of the fieldwork period, I had arranged sacred by relIg . ious it , m uc h as the p rofane' IS' , transformed
belief d. ceremony into the
(with the permission of my research subjects) to introduce them to a local I and two coll eagues an
social worker who was to be employed as an outreach worker, covering the a comparative ethnography(Nofeil M cKe ,tf�ney and DI' C!, Fonkert) conducte
same prostitution sites that I had covered during ethnographic fieldwork. by one or another of us over aeight dI erent therapeutIc comm unities studiedd
vI' d ual studIes' peno ' d 0 some ten years . 13ecau
The introductions were accomplished but the planned outreach post was involved the colIect'Ion 0ffSImI " lar (part'IClpan . se all the indi-
'frozen' (along with other local authority posts) owing to a local authority on tI1e same general topic (the
da ta for a single comparativtreatm;nt p��cess), It, was possible to reuse those
t obs ervation) data
budgetary crisis associated with non-payment of the poll tax. Nevertheless,
of qualitative m ethods, namelye �t c aVOIde , d the us
the commitment to a male prostitution outreach service had been made
and the establishment of such a service was merely postponed, taking place depth . As a result We were a tha � b :e � tl �� 0 cov
ual c nstr
erage must be sacn�ficedaint
at a later date, contrastm , g therape ' ble t0 compare pract'Ice acro , ss a WIde for
'
,
. rang
psycI1Iat
No claim is being made here that this ethnographic research made a utic c om m ' e of
nc units (studied by MUl1lty settmgs - two contrastm ' g re
.

a Camph�1I Rudolf Steiner side ntial

contribution of any importance to national policy debates about HIV / AIDS for children with learning diffi K
services. I did participate in a national colloquium organized to discuss Users (Fonkert) a 'foster fam 1'l C U � ti :;
Y care faClh
'
��� � gan ey), a concept house' forsch
" ty oor), tWo co d
ool
rug
outreach services for men who have sex with men, but by this time a number '
of outreach services for men who have sex with men were already underway
h ouses for distu rbed adolescents (Blo or)d a(Bl nd a psychla . nt �asting halfway
(Bloor) . Accounts of the r tnc
in England, funded under the Health Education Authority's MESMAC (Bloor et al., 1988). esearch meth0 s have been day hospital
supplied els ewhere
initiative (evaluated by Prout and Deverell, 1994). The research project did not
fulfil a 'state counsellor' role but, thanks to the sustained contact with research
Comparison as
subjects afforded by qualitative methods, the project was able to fulfil a dual a stim ulus to p
ractice change
research and health promotion function and to act as a demonstration project Any ethnography
for the feasibility and content of a local outreach service . ess entiall co�p ara lv
ethnographer is takingis field �
" e m approach,
of. continuing sense data th note s' �hen e or s Wh the
e IS. selecting from a cornen
" nce. Whe
SIgnIfIca o se m oments that seem t0 h'1m or ucopia
n analysing the f'Ield notes her to be of special
and com pan, ng nu the eth�ographer IS' Jux . tap
mer
When writing the ethnogr ous sim ilar and ��:1 ;r � stmg fIeld note accounts.osi
,
A
ng
nd
aCCOunts in ord er to illu aph the W IS welghl, , g vario
these comparativ strate an� develop t�e argum � us different
e J'ud gem en ts are not confIned to ethno ent m the text . Of course
are made on a Confm UIng
eVa1uative j Udgem ents graphy: similar'
. and
routine basis by all
312

31 3
 ADDRESSING SOCIAL PROBLEMS THROUGH auALiTATIVE RESEARCH
NITY
MICHAEL BL OOR
THE WIOER COMMU

making residents responsible fo;7:eepmg

thnography is account); instead I shall simply and upo� the fi�st listed practice, that of
ref ore fol lo ws tha t one possible use of edesc tion of
, reSIdents m treatment.
research subjects. It ry day comparative judgements : a rich prarip
the
ctit ioners
to assist in these eve the rap eut ic prac tice , for example, can assist serving what
particular kinds ofve judgements about their own practices, pre tion of new residents in treatment

in making evaluatiod practice and experim enting with the adopnography of All non-custodial treatment institutions fac� problems associated with the
seems to them go seems appropriate. In effect, reading an eth s and being 0
premature departure or self-discharg � f reSIdents. It is a commonplace that
practices where thisunities can be like visiting other communitie sted experi­ persons who discharge themselves ,,�:;�:0r':: completion of 'hei, " ea'ment
therapeutic comm one's everyday routines in the light of contra community :;:oy de,ive le" benefit fmm 'ho' n an those who stay to complete
drawn to reconsider s describ ed how a group visi t to a second make meal­ e course; Indeed self-dischar e rnaY be Part of a process of relapse to th;s' e.
ence: McKeganey ha 'Faswells' psychiatricocc unit to try to P athogenic patterns of behavfour tllat l ed to th� reSl' d ent's referral to the
led s taff an d reside nts at the
nit y and the rape utic asion, such as they had 1 peufIC community in the first P1:�' In ,tud,"; of 'concept hou,es' fm
hera
mu ).
times much more case at the visited community (Bloor et a1., 198y 8:of180
of a co m
tion ��
the treotmcnt of drug u'ecs, fm exom l : repm'ed sU''',,'' coks io cemain-
observed to be the practitioner audience reta ins some autonom inition of
Provided tha t the ly a minimum def
fu nc
0
:g drug -f.cee omong those ex-,"silon;, complete theIr courses have to
in its e very day w ork (ar guab can thus serve e set agamst the fact that up to three-quarters of enrollees may d'ISCharge
an d judg e me nt
tice (Fr eid son , 1970 ) , then any ethnographaly eth nographies themselves prematurely, agamst .
. staff advice ( see e.g. Volkman and Cressey's
professional prac ctice change and a number of sociolo gic son (indeed (1963 ) evaluation of the first concept h ouse, Synanon).
pra ining courses for this com rea
as a stimulus to to pro fess io nal tra unity No therapeutic commum' ty IS. more vulnerabi� to premature self-discharge
have found their no graphy is used on at least one therapeu go bmeyond
way on tic .
than a ps�chlatric day hospital like Aberd een s Ross Clinic day hospital
eth d to
our comparative wever, my fellow researchers and I wishe pra ctitioner where patIents who wish to dro' t ha e the simple expedient of failing ;0
1

training course). Ho viding opportunities for such comp arative particular �� �

turn up for treatment on the fob in ay. Treatme.nt � t the day hospital
merely passively pro ed actively to draw the attention of readers toto us, might �
was co�d�cted on a group basis and f ll d the pnne>ple of '",olity con­
judgements; we wishin one or two communities which, it seemed raphic texts
features of practice it by other communities. The utility of ethnogreaders those
: ���
mntatmn (Morrice, 1979), the " flection c to pahents, inf",molly and in
m�al group thccopy sc<sion', that thelr' conduct IS unacceptable and the
be adopted with prof can be enh a nced by making explicitrtic for
r practices � ep� chon of the therapeutic communitI' as a local� where new and less atho­
gemc social behaviours can be ex!: ented w y th and ad opted. Although
dien ces
for practitioner researcher judgements that have led to pa ula
au
silent and implic it
and ana lys ed in the first place. No authorit at
es
ive
tha
scientific
t seemed confro�'otion could be monifcM in ��
y ",ms othe, thon angry denunciation'
being recorded I s im ply list bel ow som e practic al udmg gentle imn y and h esitan' concer�, staff were aware that th
d here; dissemination. The finour
judgement is intende myself to be wo rth y of wid er
lves shared � 'd
reatment method put pressures . en
����� � wluch co�Jd [eod to ,eJf-d";char-g �
to my colleagues and lie i n whe ther practiti oners themsesfu
cces lly; successful o� even to ,uicidol impul,es. T , � emature dIscharge and self-harm
test of their utility w oul d
them se lve s able to ado pt them su
communities may notr
t er? was a convention in the d hoS�Ital (understood by staff and all bu �
judgement and foundue
adoption in the uniq
cir cu msta nces of ind ivid ual
of a clash with other valsue
d practices, o '
��
nOVIce patients alike) that feIl - patIents . should
. m provide the necessary
cau se comfort and support for patlents to re.m m treatment. In the first field note
alw ays be pos sibl e, pe rha ps
tim
be
eta blin g prob lem s. The practice we commended �
ex�ract below, 'Lenny' fled after his i rs ever p �bIic disclosure of his sexual
inadequate resourcens,oorparticular order of importance):
were as foll ows (in �
onentation; in the second 'Dawn' fle , th reatenmg suicide:
residents in treatmentthe ;
res po nsib le for kee pin g [H]e: d re�arked that he couldn't f ace tellmg . .
hIS mother he'd ha d a sexual
making fellow-residenres ts lity of
wareness of the changeabi '
2 w ays of in cre as ing ide nts ' a .
of patIents �: �
relatIOnshIp with another man . . . he d w lk d o t on Friday lunchtime and a group
had run after him and broug t 1m ack and got him talking a bit.
community structurae;way of promoting resident reflectivity;
3 the 'after-group' asresidents at staff change-over meetings;
.
4 the attendance of way of countering institutionalism;
ThIS afternoon considerable p ressure was put on 'Dawn': she had sp 0ken of her
feelings of hopelessness and depreSSIOn,. her failure to 'work' in the group, and her

5 the 'tight house' asofa participating staff; feeling that she ought t0 Ieave the day hos ital ' S everal staff members had already
. 'Edith' (stafD sar:Id She had seen Dawn glance at the clock
6 resident selection ff.
of satisfaction to junior sta
left for Pnor
l
' appomtments.
ng of alt ern ativ e sou rces several tim es.. now was her chance to end it (the roup). Her voice breaking, Dawn
7 the offeri .
IC �d up her bag, said she'd end it all right ru;hed �ut of the room. Edith did
e on all thes e poss ible m eans of improving �� ;�
ot mg to stop her. At 'Harry's' (patient) bi ng, OlIve . .
e to enl arg a funer
or et a1., 198 8 : 172-85 , for
went after her,
There is no space hernity
(patIent)
pra ctice (se e Blo
therapeutic commu 315

314
 ADDRESSING SOCIAL PROBLEMS THROUGH Q UAlITATIVE RESEARCH
NITY
THE WIDER COMMU
MICH AEL B LOO R "
n tomorrow . O�ce
other in treatment) was reported by myself and my research collaborators in
ma de her promise to come agai eed, this dashmg the usual way in the academic and practitioner press and in a paper to an
;
and
�X off and her fellow-pa urrence - dith co�ld predict that Dawn
in the toil ets . s set off. fter her Ind
caught he� hent international conference of practitioners. But we also fed back findings to
befo re s he ash ed
. IS ' ly common occ
. a faIr �

individual communities involved in the research. The feedback took various

after bolting pahents forms: in one community, Neil McKeganey and I used our knowledge of
r.
would be looked afte eturn
. ntl ' Ch ar e themselves by failing tongr th therapeutic community practice at the house to produce a video of everyday
��
Moreover, patients whold expect a � � atio� of fellow-patientsdefa
did slle visit
u
�
te rs
:�
a
practice in the community, which was subsequently used by the house for
PH. purposes; in the day hospital, I circulated to staff and to some ex·-patients
to the day h osp ita l cou
. D et ermi
· ned wo uld -be
n.
at their homes, urgm g th em to re tur
their decisions in the fo rma1 groups . or res ort (with whom I was still in contact) a research report on the relationship of the
d the
either to an no un ce an d d e fen
ans we r th e do or a nd even, in one case, leavmg patient culture to the formal treatment programme, and that report became
the basis for two 'focus group' discussions (Bloor et al., 2001) with staff and
to
to subterfuge -- failing . ow-
for k eepmg their fell ex-patients; in the two halfway houses ('Ashley' ansi 'Beeches'), I circulated
country.
This practice of makm
. gy � �a � ' en t res po nsib le
day hospital, is not fro W o�t
.
Ith a comparative report to staff of both houses which was used as the basis for
ec t ve in the
atients in treatment,Foalb
eIt e
ple, a y atie nts who stay ed away m:htus,e a focus group discussion with Ashley staff and for individual staff interviews
� re
re xa m
otential drawbacks. that they wou� ; �sited and solicited to
�
return:
ed and defaulting pa,tlent
�
with Beeches staff (who had largely dispersed to other posts in the meantime),
but also at the conclusion of the fieldwork I had previously given staff some
hospital were aw aomf dist ress
the provision of c courag�m g at�e:�on_seeking behaviour . How
ort a d sU P t o ever . suc impressionistic feedback. At Ashley, this impressionistic feedback focused
as en . ng, since staff who are a wa re of them can ralse on how one might combat premature self-discharge:
could be seen erwhe lr�\l
drawbacks are not ov rapeutIc grouys. I had previously said that I would give the staff some feedback on my thoughts
.
al the
them in the fo resident cultu re p ays an important, evenwo
rm crucial, part .m
nities: the rk of 'realIty
about the house before I left - not a 'scientific' statement, but simply an informed
The patient or .
the treatm ent procesexa s in a � K:: ��eC;:X::y practitioners as be n 1 or�
{ �
l
eb s
t era t mu
� e�0:
observer's reflections. I'd given some thought to this in advance and had decided
to concentrate on one problem I thought was perhaps inadequately attended to -
confrontation', for ducmtePd �
f llo w-p atients than by staff, wanarn. va1 s n premature departure by residents 'frightened' of the changes expected of them
effective when conplay n l. l portan art in inculcating in Thus ne .
eqU1. rmg
� ��
- and a possible solution - a stronger resident culture. I spent an evening talking
ent s
residents /pati orga11l�z tl:n n � tures and prac tice s. , � about this with the warden last night . . . and she brought it up in the staff group
understandin g 0Ide £ � � for keepin � the1r fe�l���
u� e\� �
. e resp onsi bil ity this afternoon.
that patients / res nts
a active patient / reSl e _h
� of the [t led to a lot of discussion: general agreement that the problem was there.
an exte nsio n
in treatment woalrea uld be sImp
1 r a eut ic co mm uni ties. Nevertheless, sueve At the end of the [weekly] community meeting [the warden] said that she'd like
therapeutic role dy foun� 1 ; m t
r e
P
sponsibilities, if suc cessful, could �� (after her return from holiday) a special meeting of the community to discuss the
. ts / reSI d en t s . ommumtle5. problem of people leaving.
an extenSIOn 0f luence on pa n f �elf-discharge m m any curbe
p a t '
l en
::
an appreciable inf mple, one 0 �e :;o alfway houses for distwelcome an�
d adoles
Ashley was not the only study community where the comparative analysis
At 'Ashley' for exa althou h the re�1. d ents were prepared to ge themselves acted as a spur to modifications in practice. Sociological description of every­
Cents in the stud. y,1s, establi� shed. reSIdents who chose to dischar . never day therapeutic work can act as a stimulus to practitioners to re-examine their
support new arnva . . ' ng l. hable from abs condmg) were. sta
(a procedure someti m es m � : � r fellows. In one celebrated m nce
dlst practice and perhaps modify it in response to comparative data. This stimulus
confronted or persuork aded otherwIse Y heiost the entire resident group kn�w to change can be increased by choosing certain forms of dissemination in
at the h ouse, aIm e d h ang1l1g preference to others, for example by explicit highlighting of examples of good
d un'ng my fieldwne resl' dent' s pIa nned. 'escapel , which involv . gl' ra'
. tercept the postman and appropnate 1lIS" ose
practice, and by personal briefings as well as written reports. It is also possible
b eforehand of o et outsI. de to m . one ch . ta that the close personal ties that are built up with research subjects over the
around on the stre eque) b efore tching an mter- C1'ty bus ' No possesSIOns Course of ethnographic fieldwork serve to command an interested and com­
(welfare benefits chone fellow-resl, .�ent even helped him carry his mitted audience for the fieldworker's findings. Of course, modes of research
dissuade him and dissemination that command an audience among therapeutic community
to the bus station. practitioners might not be similarly influential with other practitioner
aUdiences: this should be a matter for empirical experimentation.
Feedin g ba ck
fin din gs
' 11 might be profitabl y
ut� c co �
m ;m . t practi. ces wh1C
sible for keeping h
e ae
The above list of the(�mape dmg ma g !e sid ents respon
adopted elsewhere clu 317

316
 &
MICHAEL BLOOR THE WIDER COMMUNITY
ADD RESS I N G SOC IAL PRO
BLEM S THRO UGH QUAL
ITATIVE RES EAR CH
anal'y�able as end eavours wh
�
CON CLUSION

�n CItIzens. In a new twist on Becker�hto�� , t dIsc.Iphnary g rip of experts

ich ti . . .
Policy influence for social researchers is quite possibly a chimera, 'a unicorn It may be argued that researche o�e sIde ar� we on?' questio
among the cedars' which is glimpsed tantalizingly from time to time but .
power, but 111 the extension of rs sho uld b :: s �
. g
ISt111 not 111 the extension of
n,
always eludes us. Some might say that policy-makers themselves are a res' IS
ference in prostitutes' street d ' tance - reSIstance to meddlesome inter-
chimera: a distinguished epidemiologist of my acquaintance claims never of patI . ents' eal111gS,
private lives . The 0pposI. teand resistance ·t0 expert orchestra
to have met one. In his younger years he would frequently meet senior .
reSIstance it provokes; researche 0 f p Ower IS. not its ab sence, buttio th
n
functionaries in the health service, but they would always claim to be merely the g�oundwork for citizen resis so the argument g?es, should be layinge
rs,
implementing policies passed down from above. As he himself grew in effectIveness of exp ert power. tance ra ther than fostenng the extension and
seniority, he came to meet the yet more senior functionaries · from above, H owever, this critical w o
but they, too, claimed merely to be implementing policies passed down new version of an old sonvie f sociolof�'cal l11 . �
g, the Son of e Lemmst �ence on practitioners is a
from above. Still his seniority grew, but still he encountered only policy
arned t�e I,essons of Ijistor
always knows best, haVing le vanguard party which
implementers. He searched in vain for the fountainheads of health service
where there was none before or tute; m�y welc0n:e the provision of a service
this critical view, that male prosti y. It matters not, in
policy until in old age the truth struck him that no one knowingly makes
policy; for reasons perhaps of protective coloration everyone is convinced munities may welcome the ch� th a a lents/reSId ents in therapeutic com­
that they are mere policy-implementers, simply interpreting and elaborating by p roviding comfort and sup : t��:� ; f a � ler part in the treatment process
/i
edicts passed down from some more august authority. Analogous, if less to experts' diSciplinary po . IJ t :Ir e ows. What
colourful, arguments have been constructed by some empirical researchers the all-seeing expert: the anwer aradoXIcally, the cn. t'lca1 matt ers is resistance
of policy processes (Manning, 1989; Rock, 1987), namely that policy is a (the outreach workers and the aly�t c1alms t? know better tha yst has become
. anal
erapeutIc CommunitY staff) n the practitioners
situated discourse, a set of tacit assumptions and implicit meanings found ' s to know better thanlltho
a1so cI alm . , but the analyst
within particular offices and occupational groupings. (the prostitutes and the patie se ':h ose reSIstance should be stiffened
It is this policy discourse, this amalgam of committee asides, gossip and ar: themseIves experts, what nts / reSId ents) . ' Yet I'f tlle cnt. Ical
.
.
analysts
unspoken assumptions, that Bulmer would seek to influence through the WIth their audience? ShOUld not kin d o f dis '
cIplmary :elatlonship do they
have
gentle diffusion of ideas and research findings. But social researchers are rare can skirt these sophistries . wh the . .Y too be reSIsted ? It follo
ws that We
of pr�CtI. tIOn
. ' ere CItIz themselves commen
visitors to these corridors and committee rooms: their capacity for cultural ers, then it is not the placeens 0f researchers to murmur of rk d the wo
diffusion is minimal. The argument in this chapter has been that the real conSCIOusness and demand reSIS.
opportunities for social research influence lie closer to the coalface than they below) t ance to pastoral care. (See the summary false
do to head office, that the real opportunities for influence lie in relations with table
practitioners, not with the managers of practice.
This role for qualitative researchers as practitioner helpmeets will not be Unsustainable su gge
sted roles for
found by some to be wholly satisfactory. All practitioner-client relationships �searchers Pro blems
(be they outreach worker-prostitute relationships, or therapeutic community
staff-resident/patient relationships) are power relationships. In a Foucauldian Researchers as 'socia
l engi n eers' " Low tech n ical comp
analysis (see e.g. Foucault, 1980), power cannot be wished or legislated away, .. Lack of inst itutiona
eten ce
it is inherent in all relationships. Therapeutic advance has as its corollary Research ers as po licy l authority
form u lators/ .. P olicy- make rs rare
the extension of the controlling therapeutic gaze: the growth of public health evalUators ly seek po licies
from researchers
medicine since the nineteenth century, for example, has brought great .. P olicy-makers may
health benefits, but it has also subjected populations to increasing surveillance co m m ISS
researc h on pol icies i n
.
' lon
and regulation (Armstrong, 1983). Surveillance as a technique of power ('the .
eva lua tion
orde r to delay
eye that knows and decides, the eye that governs' - Foucault, 1973: 89) is Re�earch ers as 'enl
increasingly complemented in the late twentieth century by other techniques, pol icy- m akers
ighteners' of
�
R esearc ers h ave
few oppo rtunit ies to
most notably that of 'pastoral care' (Foucault, 1981), whereby clients of ming le �� th policy-ma
to practition ers
kers , as opposed
agencies find themselves 'shepherded' in disciplinary relationships with prac­ Research ers as
emaniCi pators
titioners whose avowed goals are merely those of care and advice. Assisting I n�olv�s a p rior com
m itment to an
in the extension of outreach work to new populations, or suggesting ways to objective rea lity reveale
d to the
increase the effectiveness of therapeutic community practice, are each alike ____ rese r h e r b u t n ot
______________________ _ Jpra
_ __ ��� ctition
�
,
� �?��ers
���or
��laity
���
__ _
__ � �=�=��__
to Po1 iCy - m akers
,

318

319
 I

ADDRESSING SOCIAL PROBLEMS THROUGH QUALITATIVE RESEARCH

' rela�lO. �ship, based on lengthy

MICHAEL BLO OR THE WIDER COMMUNITY

should seek utto the assist the research findings, A close personal and workmg
(of whe the r or not soci al rese arch episte­ social contact and built up over weeks and;onths, .IS lIkely to ensure that not
This issueclients and patients) is part of a broader debate abo y can and nly will practitioner research sub " ects ha e a partIcular mterest in the find-
resistance of us of social research, about whether value neut claim that ralit � (because of the identity of the Jresearc
mgs h.er. as much as a partlcu1ar interest
mological statain a constitutive principle of social research. Them two sides. m. ti.1e research topic), but also the ma�.be ':Illmg to devote an unusual amount
.

should rem can and should be value neutral is under attack frosearch should of tIme and effort to discussions � tl e d m s. At. t�e da� hospital and Ashley
social research joined by thombr
tle has beenrtici se who argue thculaat rrepolitical aims, halfway house, staff (and a group o} e��d y�lospltal �atIents) were willing to
On one sidetly, bat tically pa patr patory, e acin g parti , battle has been read a quite lengthy research repor.t and t:. en make theIr reactions to the report
be explici mbapoli the
or iarchy.cticOne or poli other side riptions can be the basis for a special focus group dISCUSSlOn Where the. res,earc11er has become,
such as co osetingwhoracarism that prastances, sincecyall presc
no for research subjects, a person for whom th�Y have a speCIal . th atregard
'

as a result
joined by thresearchers ungue der any circ um knowledge issusoci
m
ally
peri or of long familiarity then it s h ou Id
. have 'interest in in1P1 ementmgconle as no su .
rpnse those research
offered by and there are no grounds for the researcher to clai su. bJ.ects WIll . the researcher' s suggestions
' may become a partonof
constructe d
most clearlys changes m. practice, In effect the ualitativ e researcher
knowled ge.ument about participatory research is perhaps seen(1990
The arg ses which greeted the publication of Foster's ster ) finding his or her local practitione� .col�ctiVI. ty an� trades on that position as a
in the resp of evidence for racist practices in British schools. s or thatfouthey
on Fo nd collectivity member to disse' mmate research fmdings . . the
on the lack ce that black pupils were treated unfairly in les ed the evidence son In. re�pect of other practitioners (who are not research subjects) ,

'
little eviden cated to ability groupS; moreover, he re-examin flawed. The qualItatIve
.nch descriptions researcher has the advantage that the research methods allow
everyday prac
. . to J·ofuxtapose
.ImagmatIvely . t·Ice wh'Ich enable prcl. c't ·t·lOner 1
.
' audiences
were misallo fou nd in earlier stud ies a nd foun d it methodologica lly
itted to somye . . There is thereforethan elr own everyday . for pprac f
rac Ices WIth the research
of racism ated a considerable critical response from those co rove mm
ner
study ge ti-racism. Hammersley's (1995) review of the cont rsy firml descnpbon.
. judgements about thelr. own practIces
atIve
opportumty . and expenment .t · tlOners
1 · . to make evalu-
.

WIth the. adoption

version ter's position against various implicit and explici fromnota
of an t cha rges, bly of. new approaches described in the research findings ' Q ua 11 · tabve studies
Fos
supports dle-class white male he was experientially disa the primary ble d coll ect­ , . '
that as a miderstanding evidence of institutional racism, and thatthe changing of everyday practice offer sufficiently
as a spur to judgement and extrer�. mentatlOn . IfdSchon
detailed escnptlOns of practice to act
(1983) is correct in his
ing and und research is not the production of knowledge but
objective of argument that professional w k mvolves th·e deploym 0
. e.nt research f knowledge-in-
of society. ument of the 'strict constructivists' (the term is B est's,for1 989) that acfIOn rather than scientific knowledge, then qualItatIve allows
arg ba sis claiming professional practitioners to reflect uPon that, .previously taken-for-granted'
The
researchers owl should be was silent on social problems (havinarg noand Pawluch (1985). knowledge-in-action. ReIatedly, where speCIa 1"ISt servIces · (such as male
) statedciallsucc oolg
inctly by Wludi r) relates somr ea prostitute outreach services) do no t currently exist qua 1
l·t · research can
atIve
superior kned voluedge nd espe y his own conc ng ch apte proVI. de detailed descriptions. of1 tl1� CIrcu . ll:stances and ' behaviour of potential
Best's edit nses tomeW(aoolgar and Pawluch's paper. He argues thestru case fo na aSSIstance IS giv · th h d
re spo
of the constructivist' position in distinction to the 'str ict con ctivist' service-users such that mate
. In the special case of outreach serVIces . ethnographIC· fIeldwoof�targeted
en WI t e esign
'contextualest is unclear whether it is practically possible to achieve the strict servIces. .
so many SImilarities with outreach work that. a suc ' �essful ethnographIckproject shares
posi tion . B wh olly free of assu mpti ons abo ut obje ctive can act as a feasibility study or 1dem nst atlOn project for an outreach service
constructivistess' goa l of anal yses assu mpt ions may creep in at
reality; he citr ofvario us exam ples of how such sts, in contrast,peting may Shaw (1996) has recently de �. ope� at �ength the argument that qualitativ�
back doo such anal yses. Co ntex tual constructivi deb atin g com methods can provide a par:. Igm or exa.mplar for practitioners seeking to
p
the
collaborate wit h collectivity members in examining and reflect upon
. . and modify thelr work practIces
) r��btlOner autonomy is variable in it� extenSIveness . but universal.
policy claim s. refore, that qualitative researchers may address socipract al problems I ractIhoners may not always have the 1oca1 autonomy to deveI op new services
It seems, thecan address them most effectively by influencing procesitiosesnerof at new target populations of clients' but. a11 practItIoners . . have the autonomy
and that they itative research has a two-fold advantage in these who are the to modify their everyday wo�k ;::���e�· In seeking tl:e c�imera of policy
practice. Qual advanta ge relates to influencing practitioners influencing infl.uence, sociologists rather e l l ow research fmdmgs can address'
influence: oneresearch subjects, and the second advantage relates togs. I n respect sOCIal problems through the encouragement of mod·fI lcatIOns " and develop-
researcher's who are the wider audience for the research findin can call upon
practitionersners who are research subjects, qualitative researchers subjects as a ments m. practitioners' e daY PractIces. . The
. effectiveness of research in
addressing
.. social proble :? 11 as b een .
mcreasmgly
. ques f lOne d and even the
of practitioexisting research relationships with their researchsponse to their 1egltImacy of social research in a ddressmg those problems has been queried.
their pre- ensuring an attentive and even sympathetic re
resource for 321

320
MICHAEL BLOOR THE WIDER COMMUNITY
AD DRES SING SOCI
AL P ROB LEM S THR
OUGH QUA lITATIVE
Of course, these questions and queries have been raised most loudly outside RES EAR CH

the research community, by pundits and politicians. However, this chapter

is concerned not with punditry or politics, but rather with those questions
and queries raised within the social research community. It is suggested that
the addressing of social problems is indeed a legitimate objective of social
research and that, although the effectiveness of social research as an agency
of social change may be somewhat limited, it is certainly not wholly ineffective.
Moreover, if the impact of social research on service providers' practices is
considered alongside the impact on formal policy, then social research clearly
has the potential to be more effective yet as an agency of change. (See the
summary table below)
Qualitative researchers addressing social problems through influence on
practitioners' practices

" Qualitative researchers can capitalize .. Assisting practitioners in improving

on fieldwork relationships with service delivery may be viewed as
practitioners to stimulate i nterest in conspiring with experts against the laity
their findings
.. The rich descriptions of everyday .. Researchers should be silent on social
practice found in qualitative research problems having no basis for superior
allow practitioners to compare their knowledge
own practices with those reported
in the research
.. New practices can be adopted from .. The scope for successful changes in
research descriptions practice is frequently overestimated RE FE R E N CES
.. Ethnographies may even provide a .. Practitioner autonomy is limited ,
partial model for new outreach especially in the creation of new Advisory Coun cil O
� is
n the Mis
services services London: Department o f use of Dru s 1 988) -:IDS and Drug Misuse:
Health and SO ia ecu Part One.
' I Anatomy of the
Armstrong, D. (1983) nty.
P0II'tIca
ACKNOWLEDGE M E NTS
,
Twentieth Century. Cambrid
B ecker, H. (1 967) 'W
� �:
B d Me ical
ge: Cambridge mve � Knowledge in Britain in the
rslty Press .
oC/aI Problems, 14' 239-
hose s1'd e are we on?. ', S .
ages oJ,ffSSues: Typify '
Best, I· ( ed.) (1989) 1m
I wish to thank Ian Shaw, Anssi Perakylii and David Silverman for their help­ mg Contemporary Socia
"
48 .
�
NY: Aldine de Gru l Problems. Hawth
ful comments on an earlier draft of this chapter. All the research reported yter.
0,::
�. , M ., M cKeganey,
on here was supported by the Medical Research Council. I wish to thank of a Rang of Therape
�
N. and Fonkert, D (1
utic Com mun ity Pr�ct '
988) .
Foot m Eden: A SOCiolog
Ice. Lon on: Routle
Marina Barnard, Andrew Finlay and Neil McKeganey for their help as my Blo�r, M ., Fmlay, dge.
ical StUdy

� ��:
A, Barnard M d M
co-fieldworkers in the male prostitution study and I wish to thank Dick nsks of HIV infection in � � 7'
gan N. (199 1 ) 'Male
pro
Fonkert and Neil McKeganey for their help in the comparative analysis of Bloor, M ., McKeg
GI sg
aney N ., Fmlay, A . an
Fin port , ANS WER, A21 2: stitution and
1 -3 .
the therapeutic communities data. :::r
of s�ch o-social m'
be vlour among
d Barnard M
odels of risk behaviour ' ' (1992) 'T.he inappropriateness
to u d erstandmg HI
Bloor, ., Barnard,
Glasgow male prostitute
M., Finlay A and M K ' �
s' A1D care, 4: 131 -7
.
V-related risk

pro�ti�utes: R fr ng c
� :��
among Gla sgow male ey, N. (1 993) 'HIV-rela
NOTES
ted risk p ractices
AnthroPology Quarte oncepts of risk behavi
Bloor, M., Fra rly, 7 : 1-1 9. our', Medical
1 . This therapeutic community has since been relocated from the Ross Clinic to
nkland, J., Th omas, M
Research. Lond on: . an d RObson, K. (200
I
Sage. 1 ) Focus Groups in SOCla
Kingseat Hospital, near Aberdeen. Bulmer, M .
( 1 982) The Uses of Soci
.

al Research. London . .
' AIlen and UnWm.
322

323
MICHAEL BLOOR THE WIDER COMMUNITY

Carey, J. (1975) Sociology and Public Affairs: the Chicago School. London: Sage.
Davies, P. and Feldman, R. (1991) 'Male sex workers in South Wales', Project Sigma
Working Paper No. 35. Colchester: University of Essex.
Foster, P. (1990) Policy and Practice in Multicultural and Antiracist Education. London:
Routledge.
Foucault, M. (1973) The Birth of the Clinic, trans. A Sheridan. London: Tavistock.
Foucault, M. (1980) 'The eye of power', in Power/Knowledge: Selected Interviews and Other
Writings 1972-1 977, ed. C. Gordon. Brighton: Harvester.
Foucault, M. (198 1 ) 'Omnes et singulatim: Towards a criticism of political reason', in
Refle ct ion s on
S. McMurrin (ed.), The Tanner Lectures on Human Values II. Salt Lake City: University
of Utah Press.
Freidson, E. (1970) Profession ofMedicine. New York: Dodds Mead. Gale MtJ/er, Robert Din
Hammersley, M. (1995) The Politics of Social Research. London: Sage. g wall an d
Hammersley, M. and Atkinson, P. (1 995) Ethnography: Principles in Practice, 2nd edition. Elizab eth Murphy
London: Routledge.
Janowitz, M. (1972) Sociological Models and Social Policy. Morristown, NJ: General
Earning Systems.
McKeganey, N. and Barnard, M. (1 992) AIDS, Drugs and Sexual Risk: Lives in the Balance.
Milton Keynes: Open University Press.
Manning, P. (1989) 'Studying policies in the field', in J. Gubrium and D. Silverman There are grow'Ing d emand . . ies
(eds), The Politics of Field Research: Sociology Beyond Enlightenment. London: Sage. organizational governance ansd In contemp orary SOCIet .
for transparent
pp. 213-35. ciency, effectiveness, equity and � K ;ss��n,ce based on evidence of effi­
u l't
Morrice, J.K. (1979) 'Basic concepts, a critical review', in R Hinshelwood and number of practical and ide olo i� i Y ' ese demands are related to a
N. Manning (eds), Therapeutic Communities: Reflections and Progress. London: operations of such public and �� .gIc': a1 chanl?es .affecting the structures an d
�: �
Routledge & Kegan Paul. pp. 94-11 l . hospitals, factories, criminal organ�zatIons as corporations, sch
Prout, A and Deverell, K . (1994) 'MESMAC: Working with diversity - building
The changes include the reorganiz f e�CleS a�d �ocial service agenc ies .
i ag ools,
communities: An evaluation of a community development approach to HIV
technologies (particularly l'nform�a I.on 0 orgal1Iza.tIOnal work around ne w
prevention for men who have sex with men'. London: Health Education Authority. atIOn technologles,), Im . p 1ementation of
Rhodes, T., Hartnoll, R, Johnson, A, Holland, J. and Jones, S. (1991) 'Out of the agency re-engIneering pr ogrammes In
c
.
. bUSI . . ss
and on to the streets: A review of HIV outreach health education in Europe and the (Hammer and Champy 199 3) con ness an d n on-bUSIne . g of organizations
creati. on and learning wI " cern for better managIn
United States', ISDD Research Monograph No. 2 . London: Institute for the Study of ' 111 organIZa . ns (C
haW1a and Renesch 199ge
knowled
Drug Dependence. Wenger et al. " 2002) and thth' ' tIO
e emergence of new management 5,'
SIZ . the
. Ing '
Rock, P. (1987) A View From the Shadows: Policy Making in the Solicitor General's Office. complex paradoxica1, In . teracbv . theories em
. Ing
. fea ph a-
Oxford: Oxford University Press. of organIza ' tional pro' cesses (Stacey et al e and self-organIz
Schon, D. (1983) The Reflective Practitioner. London: Temple Smith. In this environment, organizatio ;
200 0 '
d
S
:
:e
�
.
tfIe�d, 2001).
tures
Schutz, A (1962) 'Commonsense and scientific interpretation of human action', 111anagers, professional staff and nal st kehol e pOh<:T makers, planners,
. orm clie
ation about outcomes andntsprocan not afford to Ignore valid sources
in Collected Papers, Vol. 1, ed. M . Natansson. The Hague: Martinus Nijhoff. of Inf _

pp. 1 7-38.
research is a useful but und eru . . ces ses. Qualitative social scientific
Shaw, 1. (1996) Evaluating in Practice. Aldershot: Ashgate. t 1l Ize d - too l f or a . g ma
ddressIn
_

Silverman, D. (2000) Doing Qualitative Research, a practical handbook. London: Sage. problems and dilemmas faCIn . g con . ny of the
Per. haps the best-known contem tem po ra ry org anIza
' tIo na l stakeholders.
Silverman, D. (2001) Interpreting Qualitative Data: Methods of Analysing Talk, Text and porary example of how qua 1 Ita
' . c research may have p ract'Ica . ns" istlve social
SCIentIfI
1 organiz a t'IOna 1 app l'
Interaction, 2nd edition. London: Sage.
Volkman, R and Cressey, D. (1963) 'Differential association and the rehabilitation of Wenger's (19 91) an
alysis of situated 1 earnIn Ica tlO Lave and
drug addicts', American Journal of Sociology, 64: 129-42. .
Lave and Wenger (1991: 65)
.
g and communI' t'Ies 0f practice.
Woolgar, S. and Pawluch, D. (1985) 'Ontological gerrymandering: The anatomy of ap prenticeships based their analYSIS. on fIv . e qu . .
alItatIve studies of
social problems explanations', Social Problems, 32: 214-27.

wives in Mexico � . . , among

among Yucatec Mayan mid
111 Liberia . . . , in the working- Vai and Gola tailors
�mong butchers in U.s. supermlearn ing setting of U.S . navy quartermasters . . . ,
arkets . . . , and among "no
111 Alcoholics
Anonymous. ndrinking alcoholics"

324
 USING QUALITATIVE DATA AND ANALYSIS
THE WIDER COMMUNITY
ELIZABET H M U R PHY
ERT DIN GW ALL AND
GALE MILLE R , ROB
. p�rspe· ctive in his study of theal. (20 social about how A was transformed into B or why an action that proved effective in
has ex ten � ed tlus
Wenger (19n98)0f med·lCal claims proce.ssmg, . M.ore recently, Wen g er et 02) one work site was ineffective in another. Quantitative researchers are not,
how . goa1 s bys
org a mz. a t'lOn however, alone in these difficulties. High-level organizational officers are also
organizatioe1 ope d the practical I. mphcahons by showing eve frequently at a loss to answer how and why questions about organizational
have dev busines . orgam. zat�. ons) might better achi theIr outcomes.
(especially Communrt� les. of practIce' informed and According to Brown and Duguid (2000t these problems are easily
Cultivating
buil d on � n bu n d t h' tory of theoretically ial sciences. explained. Outcomes researchers and high-level officers look at organizational
ies �
These studpplicable qualitatr;e :iz��ional research in themachine opera­ soc work processes from the outside. The outside perspective . often blinds
practically ample . ude Roy s (���9) r�search onindindustr ustrial observers from seeing the importance of worker improvisations and 'local'
Classic exa a1 tons' smd wor k on b usm ' ess a nd ial managlertm' s.afMore na1 knowledge in shaping outcomes. Brown and Duguid suggest that this is
(19 59)
tors and Ddies i clud Kanter' s ��77) �u:r�tative researchreauincrats a mu lOwall why high-level officers (and perhaps outcomes researchers) are prone to
recent stuon, LIp� sky�s (1980) s dY t eet-level bu abuse an, Ding d neglect place too much faith in formal responses (such a;; work Ire-engineering')
corporati 83) analysis of publIc. agency resPonses to child o Drug Stor es, Inc. to organizational problems.
et al.' s (19 ghan, s (1983) study of corporate fraud by R· evcfailures Im . pl 'lcated Brown and Duguid contrast the outside perspective with the inside
cases, Vau r (1996) examm. atlO' n 0f the organi zational perspective, which focuses on the details of workers' shared organizational
and her lateof the Space Shutt1e Challenger ' ns made knowledge and their everyday actions and interactions. This is the domain of
in the loss pter draws gener�l' lessons frOm the applied conscietribncutio qualitative organizational research. Social scientific qualitative researchers
This cha other, qu alitahve researcI:lers in the social uaties.onItresisearalson1 describe and explain the social conditions under which organizational work
by these, an d scie ntlf . lC. hter
' a on ualitative evalman, 1999). We is, and is not, done effectively. Both the descriptions and explanations answer
b y the soci al
informedmple, Patton, 2002; Shaw, ���� . S�w and Lish qualitative questions about the how and why of organizational outcomes. They also often
(for exa four related and overlappmg applied contributions of
. ' make fundamental components of organizational processes visible to outside
emphasize stakeholders like regulators or other public policy agents.
research: . ts for observing and analysing A classic example is Orr's (1996) study of the practical knowledge and
tive research ers' sta n d po m activities of technicians who service and repair Xerox copy machines. From
.. Qualitazati proce�ses. n � fOcu discovery in qualitative research. the outside, the technicians' work was straightforward, if not predetermined
organi onal ty of de�lgn a . � ��10ns to organizational complexity and by technological and organizational factors. The technicians were supposed
.. The flexibili ent .
istS onen a to travel to the customer's office, collect information about the problem at
'" Qualitative social SCI hand (the machines were programmed to provide some of this data) and
holism. alitative research IS. used to reframe organizational problems and fix the machine by following instructions found in documents produced
'" How quxes. by Xerox. For Orr, however, 'how is the work supposed to be done?' proved
parado disc uSS . ':111.
lO a less interesting question than 'how do the technicians actually get the job
e cont rib utio ns in tu rn. The � tive done?' The latter question asks what goes on inside this work process.
We shall disho cuss each oft'lvethesorga . .
atlOn a 1 re s earch diffe rs from qua ntita
also identify w quarIta arch, mzs w 11 as to note how they can be usefmed u11y Based on his observations ot and conversations with, the technicians, Orr
discovered that copy machine service and repair is a complex, contingent
outcomes-orie - rese �
nted y conS1� dermg h0w social scientifically . s inf or
combi r se rchnclu. £de
ne d . W e co b tati ve approaches to orgamzat'IOn and improvisational process. While their training emphasizes technical com­
r uali
_

fer he petence, effective copy machine technicians are also managers of knowledge
I

qualitative �m �ItS et�c 0t,�:�;nd�:lin; (DingwalL 1 992) . and relationships. The technicians must build and manage effective working
particularly relationships with the machines (many of which are distinguished by their
idiosyncrasies and seeming unpredictabilityt the customers who operate
N I ZATI O NAL P
R OC ESSES and often misuse the machines, and the larger technician community. The
I N S I D E O RG A
. usef ul for examining rela ti�n- technician community is an important source of practical knowledge about
a rch de sIg ns ar e articula rly
. zat'10nal work' They can vlde. -
pro how to respond to troublesome copy machines.
Quantitative nrese
. s an d 0utp uts
t:
m org am
put me data wh ' ch in rinciple at least, alloW adwnar The community's knowledge is 'local' and unofficial. It is often conveyed
ships betwee relimable o 'takeh10l'der� to decide whether a partie pculrob-h as stories and certainly cannot be found in Xerox's technical manuals. This
detailed and1IC· makeoutc s a nd otI1er s rt th ue shared knowledge addresses the distinctive problems posed by the idio­
Y wlu:�e. H owever, while these studies canablerepo
istrators, pow orth syncrasies of individual copy machines. 'There is an assumption on the part
action was action A w111 lead to outcome B , they are rarely to us m tell
ability that 327

326
GALE MILLER, ROBERT DINGWALL AND ELIZABETH MURPHY THE WIDER COMMUNITY
USIN G QUA lITATIVE DATA
AND ANAL YSIS
of all participants in the service world that technicians can solve any machine implementat
problem' (Orr, 1996: 71). Thus, the problems posed by a seemingly un­ its promises. ion' the hospI' taI concIuded that the plan had failed to deliver on
repairable copy machine are treated as challenges to the collective wisdom Wiener's longed immersion in the settin
and competence of the entire technician community. of the procepro
sses involved l'n the emergence dgeSIgn and car. eful documentation
Orr's study is ironic and perhaps debunking. It shows how the actual eventual kdown of the red.eSl" gn Im . .tIa. tIv. 'e allo'we,dImp lementation, and
work of servicing and repairing copy machines requires that technicians of· the facbrea
tors that turned a seemmg h er
. tot orga
'dentIfy
' some
sometimes ignore official (and authoritative) organizational instructions and change into a costly, protractefd a'n dlyd'Iratruph
IOn
' .al and. efflClen 1

procedures. Some readers might conclude that the study teaches us that the factors into two major types' d:�:. en�l. � �m theve faIlure. She classifiedatio
c

niz nal
'real' authorities on how organizational work is done are the people who plan and hospital manage;s' � �sse execution of the work redesignse the
do the work. We do not necessarily disagree. But, for us, the more important knowledge, analytical skills and p��e��t:�l or ssm�n.ts of .the medical staff's
lesson taught by Orr's study involves the way in which the technicians' latter assessments were the baSI. S for managemeadmmlstratlve leade. rship. The
improvisations and apparent rule violations advanced the company's most exp. anded responsibilities tC) nurses an nt deClS' .'lOns to aSSIgn new and
important interest, effective customer service. Orr's research is a study of the , d oth er sta
tahve analysis of the medical staff's. work shou ff, members. A close quali-
social organization of best practices in copy machine service and repair. disruption accompanying the o ed chanld have flagged up the risks of
This study documents how and why Xerox's copy machine service and of properly retraining the medi�:l ��:f t me ges �nd challenged the costs
0
repair programme is effective. Such information is vital to stakeholders who Qualitative research is not b:na�� �o;et� the�r �ew responsibilities.
might contemplate changing work processes. N o one can know for sure what �ecision making. As Brown an� dmml�trative planning and
all the consequences of a change in a work process will be: however, the i ( O )
tIOnaI managers and other stakeh �ers. ne obsel vel though, organiza­
0
chances of unintended consequences emerging increase when managers, available to them as outsl'de bservers f woedrk to balance the inf
. 0 0
ormation
. orm
policy makers and planners cannot see the existing process from the inside. obtamed from inside. processes, WI'th mf ation
Changes that seem small and limited when proposed and discussed may have
far-reaching and unexpected consequences when implemented.
Thus, another practical use of qualitative organizational research is that DISC O VER Y
it provides an inside standpoint for anticipating possible unintended con­
sequences of new policies and procedures. Consider, for example, Wiener's Wiener's stu illustrates another fea�ure of . .
(2000) work on quality assessment in US hospitals. One part of this reports to deal with dy un an tic ipa ted fac t ors or ISs ue s
qualItatIve research: its ability
. ional
an attempt to apply the concepts and practices of Ire-engineering' and 'work stake.hol.ders with information that they d d t ProVI'de orgamzat
c
an0
redesign'. The plan promised to reduce the hospital's costs while increasing QualItatIve 0 no t exp
customer satisfaction, quality of care and market share. It redistributed is a fleXl'ble, I'terahv. e process aect
researchers tosturesdypodendsigton un II ow
to be rel.evant.
'
hospital staff members' work assignments in order to reduce labour costs (by of the research. Wiener (2000an' ticl�f' �tesc ed �bPortum.he. s tha ' t arisemgin
qualItative
the
eliminating unnecessary specialist positions, such as phlebotomists), increase arose from a 'combinationofthe' aCCI'dentnandessagho w. a valuable Source ofCoudarseta
the efficiency of hospital staff in serving patients, and reduce unnecessary idle research'. Early in her research while attend' aCItY that marks unstructured
time for staff members. the
. proPatient Care Assessment Co'UnCI'1, an organmg iza
a conference sponsored by
A major unanticipated problem involved taking blood samples from Im veme professionals she was mV f
. d to obIOnservfequfuatIure
0 Ity
' assessment/
patients, a job that was previously done by the phlebotomists but was now the council.ntThes e ' Ite
meetin;s Proved to be a key Source of data.meetm. gs of
'
assigned to other staff. Problems with the new arrangement soon became a towards the of her re earcl1, S he dlsc . overed an onIm' Similarly,
recurrent theme at meetings of the troubleshooting committee. The laboratory qualI.ty assessend
me nt/ im e n etw ork where
representative reported that staff were having difficulty in drawing blood. and sought or offered pro vement profes . sio na I s d'IScussed common
advice and aSSlStance to one another. She su ble. ms pro
Since she no longer had a team of phlebotomists available, she was unable to t o the network and
ob tam' ed Im ' portant da.ta that a11owed her to asse bscnbed
offer assistance and could only advise callers to approach a doctor. 'Flexible generalizability of her find
working' had actually led to the reallocation of tasks previously carried out Were being implemented. ings to other hospItaI settm. gs where similar posslicthe ies
by relatively cheap specialized staff (phlebotomists) to more expensive T�e evolution of Wiener's study reflects
doctors. Moreover, lack of skill in drawing blood led staff to use more expen­ �hnngsdISOrcovery in qualitative research. Qualita/Ive 11e general emphaSIS. placed
sive equipment that could more easily be operated by relatively untrained and phen researchers approach set­
inexperienced people. Similar problems arose in relation to cuts in respiratory­ :-Vhat will turnomoutentoa ofbeintimpo
erest without ;:s�mmg . that t ey kno
therapy staff and the reallocation of their work. Almost two years after IS an empirical orientation focusrtanedt.onQ�� I ahv� research� the socw ialin adv
111 sci
ance
en
IsCOVenng how organizational worldcess
328

329
 USING QUALITATIVE OATA AND ANALYSIS
ABET H M U RP HY THE WIDER COMMUNITY
DINGWALL ANO ELIZ
II

GALE MILLE R , ROB ERT

s stu dy sho w s how this orientation studies of organizations. As Shaw (1999) notes, applied qualitative research
.
rgamzed and oper ate. W ·
len er' cumstances'tous by is one way to balance quantitative researchers' traditional concern for rigour
are SOCI. a11y orea to tak e a d van age
t, of fort uit ous cir with a concern to do research that is relevant. But qualitative research can
includes a newdm·anedssunexpected sources f data. In referrinunfg toortu'fortUl
0
b ut lso enhance the rigour and credibility of quantitative research. An example
developingces, , 1 owever, we d not . ea n to endorse the arch arenate,
0 III .
ISO ted
1 a IS� Kaplan and her colleagues' study of the impact of computerization on
circumstan. pressIO. n that the diSC . ovenes f qualitative rese
0 clinical laboratory services in a large teaching hospital (Kaplan, 1986, 1987;
Im cases. WhI. le the :xpectation of discovery underl'Ieh,s many
1

common,erifi researc the se Kaplan and Duchon, 1988, 1989). The researchers studied the introduction of
and unv able 'cal and a na l ytlc p rac t'IC es of qua. lita tive 'f ymg. a computer system, which replaced manual systems for managing testing, and
of the methalsod 1 Ogl
es
0 0
inc lud e
C

p roce
.

d ure s f or double-checking and ven the reporting of results to clinicians, in nine laboratories.
strategi . at they have The researchers used a combination of qualitative and quantitative
discoveries ers use checkpom. t s m. their designs to assess wh For methods. The quantitative measures were designed to test pre-existing theory
Research d to ad'JUst their data co11ectIOn ' and anal ysi s in response.ectIOn . about job characteristics and satisfaction and their r,elationship to reactions
discovereddata an . are car ne . d. ou t alo ngsi de dat a col 1
.
example, . ve recoseadm·rcgh .anPrelid analysIS ana1YSIS. f data guides ongoing. decIsIOns
0
. to the new computer system. Qualitative data were gathered from a combi­
in qualitatitegies for future data minary IO The design may be adJusted, for nation of interviews, observation and open-ended questionnaire responses,
about stra test out tlle scope of acoge11ner ecf � ' n derived from one part of the carried out before, during and after the installation. The researchers compared
allzat' 10 and contrasted laboratory staff members' responses to the new computer
example, to . g by considerm. g whethen't h Ids under diffeecti 0
rent circumstalnces YSIS' systems in each of the nine laboratories. If the quantitative study had been
research settmal., 1998). This approach t q litative data coll groonundanedd ana the ory. carried out in isolation, the researchers would have concluded that differ­
(Dingwall etser and Strauss, s (1 967\ p�op�:als for building a chapter about ences in evaluation of the new system were unrelated to job characteristics.
echoes Glaforgotten, indeed, that G �ser and StrausS includedns of qualitative However, the qualitative study both raised doubts about the validity of these
It is often qualitative data to deve op grounded applicatio findings and suggested a means of resolution.
the use of (see also Patton, 2002) . (including Revisiting the qualitative data, Kaplan was able to identify variation among
knowledge e researchers also use muIf. IpIe research methodschecki and the laboratory technicians, which was not reflected in the quantitative
Qualitativ
. d s
0 ) and i nfor ma nt s m c olle cti ng ' double- Inngdsco, a measures of job characteristics and job satisfaction. The variation did not
quantitatIVeelr.metdISCh A n exa mp .
1e IS K anter's (19
77) study of relate to the objective nature of the work but to the ways in which the tech­
' OV eries .
verifying thal corporatio. n. From the °utset, her research goals wntereof structural to pro d uce
multination phY that would both advance the developmepractIc. aI recom- nicians understood their jobs. Kaplan and Duchon (1988: 49) described these
an ethnondm gr � and produce porate managers differences:
underst� fogsr chofang orgamzatiould
C
nal b ehaviourto high -level cor te workers and
at wo .be, usef ul
mend. atlonsve the careere thopp
•

ortumties for' lowander-lrecolmme eve corpora . are b ased One group saw their jobs in terms of producing results reports, the other in terms
and Impro anter reports that her f '
m d mgs ndatIOns . wS WI' th of the laboratory bench work necessary to produce those results reports. The group
managers. dat K
fro m sale s wor k ers an d sale s m anagers, interofvIecorporate who saw its job in terms of bench work was oriented towards the work of producing
on survey f Iandsco employees and managers, examinatio
20 5 n
onne1), gr. oup
lab results, whereas the group who viewed its work in terms of reporting results
0
a variety (m. c1udm' g performance eva' 1uations of Indsco pers was oriented towards the outcomes of the lab work: the members of this group
ngs a nd tra mmg .
documents ervat'IOn f corp 0 orat e me eti saw themselves as providing a service.

discussIO. n�, ally

and participa(19nt77obs
, Kan ter not es th at she double-checked her fmdmgs '
This distinction became central to the subsequent analysis of both qualitative
sessions. Fm .
' 296)
and quantitative data. The difference in work orientation was associated with
with different evaluations of the impact of the computer system. Those who saw
. �
le wIth whor I b uilt
hips over the
ced to see their work primarily in terms of producing results tended to view the new
close working relations
A small group of peop largely . uncflOns where they were well pla tell e system as having a negative impact by increasing workload and interfering
m
a1so
wer e coul d
1 e 111 a large num
le
Years. These
peop I
b er 0f Ievels at ndsco.
They
1 with established ways of working. The others, who defined their task in terms
111
.
er of peo p eers. couId
of reporting results, were more likely to offer favourable assessments. They
a large n umb . in their own car
rmation ab out the issues
s well as provide info
�se them to check out sto
ries I gath ered elsewhe re.
saw the new system as a means of improving the service they offered.
sco sh OW S ho w rese arch ers som e t ·
Imes incorpoenratee- The qualitative work's flexibility allowed the researchers to discover local
differences in job orientation that had more explanatory power than the
K ante
. r's study ofchInd
withi n qua l '
Ita t '
Ive s tudi also practicalIbative
es· There areents ft quantitative indicators derived in advance from general theories. The existence
'
tIVe re sea r
quantita ding discovery-on. ented quanat
nclu I ive compon in quan
fits to i 331

330
 q

USING QUALITATIVE DATA AND ANALYSIS

BETH M U R PHY THE WIDER COMMUNITY
DINGWALL AND ELIZA
,\

GALE MILLE R , ROBERT

Ies
' for managerthes dealing, with organizational problems, the findings can stimulate stakeholder discussions
orie ntat ,
IOns sug ge s t s different stra teg n tech , nS'
nICIa about the better management of ongoing problems and dilemmas, Lipsky's
of these nces f th� change' partIcu ,1ar1y a n eed to focu s o
the consequengs of then work rath er' tl1an
0
on attitudinal measures at the pom' t (1980) study of street-level bureaucrats exemplifies such a contribution,
Another example is Vaughan'S (1983) study of corporate misconduct by
a
underst t,ndi Revco Drug Stores, Inc, The company was found to have falsely billed the state
of recruitmen of Ohio for prescriptions that were neither ordered by physicians nor delivered
to patients, Part of Vaughan's study examined the organizational context that
made it possible to commit the fraudulent acts. She showed how this related
H O LIS M
re 0 f a 11 hum an activity istext a central issue are for to Revco's growth and development, and the consequent leakage 'of authority
The com plex and cont extu al natu 1 I work, con ual fact ors, n, s within the company, Authority leakage occurs when members of upper-level
organizationalberepote searc1 In mthre uch quan , t'tal've nts to a re sear ch desIg management no longer know or control what is happening in organizational
con tan� , ina
lly atenI'dmgand reliable relationshIp' s, whIC' I bY
consid,ered toe a " m IS tont'Iaesta
1,

1 subunits, Vaughan explains that leakage is not in and of itself a sign of organ­
111 va I
integnty, Thold �IrrespectI, ve, bI'ofISco t e t Suc h rese arch ers are sen " e t the
sItIv 0
present
izational problems: it is a rational response to the changing circumstances
definition hposed by the complex�Ity �i the social world and the ever-text and
faced by successful organizations.
Vaughan's advice for stakeholders wishing to minimize the negative
diffi cultie s '
IOn s" Then , ,
ann '
IS, 0
h wever toMishl
, nulli fy the con
risk of spu rio us corr e 1 a t ,
tIOn a 1 (St ak e, er (197 9) de scn 'bes potential of authority leakage focuses on two general issues, First, she identifies
totryto eliminasatet mer
he � � , lY Itua 1995) .
, ive four aspects of interorganizational transactions that create opportunities
this pra ctic e text-strlppzng
conte xt are p1 a ce d at the cent re of qua lItat ' for corporate misconduct, Second, she distinguishes between widespread
con
By contrast, , , com p 1 e XI 't y an
. d , ZatIO' ns. ('M' ller and Dingwall, 1997) , Cof 1lOW
1
ntext authority leakage in large, complex organizations and the more narrowly
socia l sci entl fIcr ese arc h 0 n orga nI study 0 dispersed opportunities for illegal activity by their members. Only some
stre ssed , not stn, pped ConSIde ' r " f or exa mpIe, Bosk's by(19J'79) u ni or house, staff organization members (two in the case of Revco) both have the necessary skills
is ego '
nze, an d san ctI On me dI ' cal erro r '
surgeons detect, cat
en tifi es seve ra 1 sl'o , rtc orn ing s in the way tha, t JUnhIor and occupy organizational positions that provide them with opportunities
in trainingeval , Bosk and h eld accountable, particularly the w�y m w h
'd 1 '
IC to engage in undetected illegalities, While Vaughan does not offer remedies to
staff were titutuate d re l '
I an t 0 n surge ons' indi vidu al Judgements the problem of organizational misconduct, her holistic approach helps her
ns are overl y
medical ins s,ioThe cred'b� TtY fhIs, recommendatI'ons for change (andeta d tho, se to identify factors relevant to organizational stakeholders attempting to antici­
11 0
and conscience nent ed re se � rc h ers ) ' how eve r ' rests on his ctiIled pate and respond to potential unintended consequences of organizational
made by othicherknowledge of the comp ex organizational settings and pra ces,
poli cy-o growth,
Marlaire's (1990) conversation analytic study of bias in educational testing
ethnograp
, na 1s mo re acco is also instructive. Although controversial, the testing of children's intellectual
untab le
fesslO
Any prog ram m atic chan ge wh'1Ch ' ntends to make pro enological understanding
� a �omplex phenom abilities and achievement is widely practised in US schools, These tests have
to clients must of necessity
start WIth
b 1l1' tY and ho' w it is achieved, Field ,
research
0 f h ow
practical importance for teachers, school administrators and counsellors who
of what curren tly pass , es 0r acco unta
rICY by ground mg 1 111
f
' 't ' a firm understan dmg
s�and-
use the scores to assign children to classes and programmes that are judged
such as this info rms po , CIal , worlds It is only from this co ncret e u nder
to suit their abilities. Educators rely on the findings despite substantial public
Partic ipan ts co nstr uct theIr
,
SO
1 der that
'
any ch anges in the
,
exis ting inte ractional
debate about the tests' potential cultural biases,
al control can be negofIat '
a or
ing of the present, pra ctlc
ed (Bosk 1979: 6) Marlaire analysed video tapes and transcripts of ten educational assess­
P oliti cs of soci
ncrete undeinsirstand' mg ments conducted by special education clinicians, Interaction between the
on b asm , g org n ' f al olicy on 'co de per­ clinicians and the children was central to the testing process: the clinicians
Bosk's emp practI, cal order� e��o:;ou� earlier discussion of the
hasi s asked questions and made statements to which the children responded, The
of the, pre However, his study SIlOWS h w, this can also sometime e soIutbe' n5st
sent, 0
s be
10 clinicians then evaluated the children's responses as appropriate or inappro­
spechves, iC, h t0 see
m Wh ' the "
'bI g pIc tur e , '
m thi s instance possl, ' b1 priate, Marlaire's data show that the issue of bias in testing is more complex
place fro lem of profeSSI, Onal g�vernance and accountability al contexts than public debates suggest. Indeed, she concludes that bias is unavoidable
to the prob ve research can clari y th e socla' 1, cultural and structur (1980: 193) in such interactions, This arises from an invisible collaboration between
Qualitati orgamza IO na1 problems and d'l I e mm
, ky
as ' As Lipsdesi children and clinicians, a collaboration that is essential if they are to interact
with '
associated is a necessary and meV f
' I' t able tension bet ween the reredtouhaciblvee effectively,
states" 'There he short run, and the recogm't', n that problems are notn'f']CatIO' n5
10 This study illustrates how conversation analytic research investigates the
an impact In t mc, rem en ta 1 ma , atI' Ons, W h'lI e the contextual cla tIOnS to
mpul present, practical order stressed by Bosk. More precisely, Marlaire shows how
to short- qualitative researchers d not always lead to easy solu
term 0
,

provi'ded by 333

332
 j

USING QUALITATIVE DATA AND ANALYSIS

THE WIDER COMMUNITY
E LIZABETH MUR PHY
ERT D I N GWALL AND
GALE MILLER, ROB
contingent sim (Rawls, 1987) , therapists (Fisch et al 1 982) F '1y �h:rapists base their work on the belief
stin g is com ple x, collab orati ve and plifies the that family member�' conti�u:: ac m ways that sustain their proble�s
educational te ators and theiandr criti cs over sting situa­ because they lack alternative languages for formulating what is ·possible in
Context-strippi ng by bothed ebducy b oth ch ildr en clin icia ns in te phrased their lives, Reframing introduces new 1anguages, for descnbmg " and orienting
interpretivnot s disp lay
e skill xample, that the clinicians' questions were often correct to problematic conditions, The roble sustammg ' patterns noted by family
tions, She chiles,dre for e al information about what might count as the�apists resemble the persist�nt PU�rIC debates about bias in educational
to give the'Thus, deci n minim g the formula for a correct answer is no simple testmg, As Marlaire (1990) notes, tI1ese rest .on, the a �sumptIOn, ' shared by
answers, the chilphe d
rin
mu st ofte n "sec o nd guess" clinicians' intentions' ProfessIOnal
' educators and their critics th at It IS pOSSIble to hav . e unbiased
matter and1990: 254) , at all testing tests, She reframes the debate, and pomts " to new , ? , P IOns,
t' b y calling attention
(Marlaire, gh Marlaire (1990: 257) might have concludedshethack nowledges f'
to t11e collaborative organization of tes mg sItuatIOns . ht randomly
Althou intellectual abilities should be eliminated, to cea se using 0
Ker Muir's (1977) study f h ow twenty -elg . selected police
s
of children'ational context and importance: 'Even if we wer e ' I ' "
' responded to volatIle po Ice-CItIzen encou�t:rs IS' a further example
offlcers
its organiz ssessment tools , ' , , we w ould still have the prob lem of how to of how qualitative researchers reframe orgaIllzahonal problems and para­
particular a ct children for appropriate program slots,' Basenizing educatictical
d o n this pra '

doxes, He draws upon both IT1 lca1 theory and his qualitative data in
c

sort and selerlaire makes several recommendations for reorgaabilities displaye onal reframing police-citizen encoE�ters as 1�gal extortionate transactions. He
reality, Maeractions to assess more comprehensively the ors affecting thed explains that both legal and I' IIega l extortIOnate
, , relationships , th's transactIOns' involve antag-
testing int and to take better account of interactional fact OIllStlC
, in Whl ' c h one party (m , ,1 c � se: the polIce) ' explicitly
by childrenscores, "
or Imp1lCltly threatens to take a hosta ge (the clhz�n s, lIberty) if the requested
children' S �ISansom (citizen acquiescence with the porIce offIcer s requests or demands)
not paid, , exercise of police
Effe�tive extortionate transactions (that ' s, th,e ef�echve
R E F RA M I N G
s ex perie nce org an izati onal complexitytheis authonty) rest on two related cond'fIOns, �he fIrst IS that citizens value the
eho l der
ch stak sometimes seem to be irresolvable, This is hostage sufficiently to be WI' II'mg to barga'm WI'th the polIce ' to keep it. The
1

One w ay alinpawhi s, which (2001 ) b ook about The Paradox of Control second is that citizens are able to paY th, e requeste� ransom, lIn the absence
as practic eme ofoxe
rad
atfield's g-term corporate manager, states that his of either hostage or ransom the e to�tIOnate relatIOnship will break down'
central th ations, StrStre
eatfield, a lon (Ker Muir, 1977: 38), When �he ext,ortIonate transaction breaks down' police
experiences have taught him that officers and citizens are faced WIth a paradox, havmg ' profound practical
in Organiz
dox of being "in control"
and "not , lIcatIOns,
Imp ' , not the least of wh' h IS' tIlat offIcers sometimes then assert
managers find that they
have to live with the para
y to live with paradox,
the cou rage to control by illegal and brutal mean��
in control" simultaneo
usly , It is this cap acit
" that constitutes Breakdown can occur in four Ircu ' n:s�anc�s: police encounters with very
continue to particip ate c reat ively in spite of "not being in control
(Streatfield 2001 : 1 40) poor citizens who have few resou�ces WIt , whIch to bargain; highly emotional
effective m anagement.
Mu ir (1977) calls mutually
family disputes ' crowds intent on protectmg their h onour �nd achieving their
The paradtory is an exam ple of wh at Ke r
ox of controlThe assertions that managers are in control andernot in go�I�; and J"uvenile activities oriented towards provmg ' one s manhood andl or
hs. sions' (K uthirs
Mu gammg prestige within a group' Pol'Ice threats of phYSIca ' 1 force and arrest take
contradic conttrut ry, yet ' each is a pre mise for man y deci
radicto the study of true but mutually contradictory tr s on a d'fferent significance l'n these SI" tuatIOns ' con, 'pared WIth ' encounters that
control areH e expla ins that for social scientists, It is a curse when other s�fISfy all the conditions necessar for the extortIOnate transaction, In volatile
1

1 977: 283), g bour SItuations threatening behavI'our mcreases y 0

the p tent'la1 for VIOlence ' and may
and a blefinssinding
is b oth a curialse scie s and anal y ses as co m mo n sense and bela that a vance the agendas of provoc a t eurs,
interpret socs, This ntif ic re spo nse whe n rese arch me
ers a ssu study d
Ke MUIr' uses his interviews with ��d observatIOns ' of, police officers
a fre quen t
the obviouprivilegeisone mutually contradictory truth oveowe e
r another. Then social to assess the practical effectiveness of elr, responses to situations where
l'

m ust
they y contradictory truths becomes a blessin g, h ver, wh rtions the, extortionate transaction breaks down notes that ten of the twenty-eight
utu all
of m focus on the social conditions under wh ich each of the asseey not of�lcers, who he classifies as Iprofessio��l�' w�re most effective in dealing
scientists example, when are managers in control and whe n are th rou gh WIth these situations, Althou;h t1;:::ferofessl�n �l, officers accepted their
is true, For? This is an empirical question that can be addressed th responsibility to arrest citizen SO i they �n,ItIally searched for other,
in control research, about difffaemringily often ,highly creative, solutions, One of ;:er MUIr s (1977: 82) interviewees
qualitative ulation of the paradox of control as a question use descnbes another officer:
The form ons is similar to the reframing techniques d by
social conditi 335

334
 USING. QUALITATIVE DATA AND ANALYSIS
THE WIDER COMMUNITY
ELIZA B ETH MUR PHY
ERT DlN G.WALL AND
@

GAL E MILLE R , ROB

. a hUSband and wife throwing and yelling
at each attempting to extinguish fires that citizens are happy to see burn or to
fam ily beef wIth he di� n't say a paramedics trying to evacuate injured people from tense crowds or riots. They
he'd come i nto a
wn on th o ch and take his hat off, and
other. Then h e'd set do couple e t ind of silly . He' d take 45 minutes
� 111 each of may also be useful to people like reception desk clerks who have to control
oner or later the �� queues in stressful environments such as airline check-ins. Finally, Ker Muir's
word. So
. ' ns, but he never had
to come back. data and analysis should be of interest to the corporate managers to whom
these SituatIO
est was nec essa ry, Professi onal offic
.
ers stIll Streatfield's (2001) book is directed. The professional officers in Ker Muir's
Even whereto the y be liev ed arr
. s�. t�ah::� . n. study illustrate one way that managers might guide and shape situations
attempted naldefu � e volatIle ym ent of con ver satlO in which they see themselves as not in control.
_.

OffIcers werople dlst thei r enjo

mg l ed b icers and citizens. In the process,
Professio talk pe e, nots-hanerdol£.needs. Professional officers also.
both
They liked tod abowith ut oth ers c1an oncer . s. K :r Mun
they learnethe impo rtan ce of
,t

guag e in olic e-ci' tizen encou nter

ed partlculathey rly
CONCLUSION

recognized ese officers as eloquent, not :ecause they possess be cau

e but anipulate se This chapter has examined the contributions that qtlalitative research based
describes thocabularies or were ad�p� at fIg.U:�tive languagmeti mes m lready in the social sciences can make to informing debates among organizational
extensive v in using language to m or� ' m ruct and so by a stakeholders about the most efficient, effective, equitable and humane means
were effective . Mui r's obs erv a tIOn s were su pported data
others . Iroby mca11 y, K e�
ol1c
.
e dep r ! ent. The depa rtm ent' s app I of theform
l' ca lOn
t'
kmd. of achieving their various goals. Whether in private or public sector organ­
izations, the prevailing expectation in contemporary societies is that these
collected que the � , you the opportu nity, and if nothing
; �
included the e comstlOmun: nity If OU rIV� ou ld y ou prefer tor,fou nd (a) a debates will take place in a rational and transparent fashion meeting legally
re
1<

existed insoci th whe 0

'
. �hestra?' (Ker Mui 7: 231). 197 imposed standards of corporate governance. Qualitative research has a
ass lca
i
deb ating offi ety. or fo ru m; (b) a c . lY more likely than the other officers particular part to play in exploring issues of process, in explaining how out­
rs w re ov e rw hel
ce ; ndm. g debatin society' . m mg comes are achieved -or not, as the case may be. Although stakeholders retain
Professional dy to cho os� fou the la�ger or ra�izational context of police-cItI. . zen responsibility for the outcomes that reflect the balance of their interests and
in h is stu dered dIdo the gpol'Ice chie£' s manageme. nt. of the bargaining strengths, qualitative evidence may help them to see new strategies
Ker Muir also asconSI ?'
encounters nbyfoster og,� ll. 'Hhlbl
m
1<.' .ow. t t�e professional orientation todpbeohcbymgpay. mg H. e and possibilities for conciliation that advance their goals.
Many different people do qualitative research in organizations for many
organizatiothat the chIef;s most Important contribution way oul
co nclu de d role of serg e � t Mos t day- to-d w ork by patronol different reasons. Kunda (1992) discusses the relationship between Lyndsville
greater attentio n � -rankm. g po�Ic: department officials. Sergeants, .ve
to th · (a high-tech corporation that he studied) and 'tech watchers'. The latter include
officers is invindsible to high posl. tlo. ned to identify effective an. d ineffectI various academic researchers, consultants, journalists and other organizational
the other haers,, asa wuellniqasuelyt� 1 the conversational techoutmqu
re
w
es used by
the po�. ce
outsiders who regularly observe and comment on aspects of Lyndsville.
Kunda (1992: 254) explains that it is not possible clearly to differentiate
patrol offic OffIc. ers. K er MUlr �:��s several suggestions ab nd,hoby ext ensIOn, between different kinds of tech watchers.
professionalmight increase the p�ofeSSl' Onalism of sergeants a
department onalism of patrol �ff�c::: stions o�fer
the professi y shows h ow socla i ntific perspectives andnsque for seemg
Published scholarly work of the "applied" sort resembles the popular genre, and
This stud guages for descnbm . . g r:roblematic situatio andxerci se. Ker
journalists often cite the popular literature in lieu of their own brand of research
alternativenses lan
foc us on lang ua re tha n a the oretical e use s to
and theory. Consequently, these forms of observation, analysis and reporting d o
new respo ns fro. The . c a d �r:�i�� qualitative data that wlth he .
not always fall into distinct categories, but belong o n a continuum.
syst e matI
Muir begi cersmW lO are eff�ct��ve �. n . ging difficult eesncothese office�s unte rs
identify offi skill lIe: s m. speclfymg Jus�:�:t it is that makys that make then In a conversation with GM, a management consultant remarked that she found
citizens. His their pe�rs and m r�f����latin g this in wa s can be offered ituseful to spend a few days observing in the organizations that hired her. At
different frome, generalIzable and e le. police sergeant them through the very least, these observations usually revealed occasions when
skills portabl1 de cnp. t'Ions and then encouraged to developnow f everyday organization members, at all levels, routinely used their time in inefficient
these generaructI�On, translatmg . them back into the here and 0 ways.
pra ct ical inst ..g There is, then, a sense in which the applied skills and services of the
.
work. er, these generalIzed accou. nts also mak e th e 1 ess on s of polIcm consultant and of the social scientific qualitative researcher are similar.
H owev others who have to negotIate 'th difficult citizense-toothfirere ecrlneWers But they are not identical. Clearly, most social scientific qualitative research
_

available to ers, for example. The same s;:�ls may b e of valu in organizations involves more than spending a few days talking with organ-
gency work 337

336
 2
GALE MILLER, ROBERT DINGWALL AND ELIZABETH MURPHY " THE WIDER COMMUNITY
USI N G QUA LITATIVE
DATA AND ANA LYS IS
ization members and observing their actions and interactions. Social scientists how they �ontribute to und esired org
are more likely than other qualitative researchers to immerse themselves in that orgamzat ani
. zaflOnal o mes. It is important
the organizations that they study by doing in-depth interviews with members, researchers. ional stakeholders recelve both messutco
ages from qualitative
carefully examining relevant records, and making extensive observations of
members' activities and/or audio or video recordings of meetings, events
or routine practice. They also bring distinctive background knowledge from
their reading of previous research reports or of theoretical literature that seeks
to generalize from an array of particular findings.
Roy's (1959) analysis of the social organization of time in a machine shop
points to a related difference between social scientific and other qualitative
approaches to organizations. The primary purpose of social scientific quali­
tative research is to understand how organizations work rather than to expose
malpractice or to advise some stakeholders on how to achieve their preferred
goals at the expense of others. In the process of understanding, of course,
important information may be uncovered relevant to either of these objectives.
Social scientists also tend to be more concerned with the general than with
the specific: even where the research only involves a single organization,
this is typically seen as a case study that stands for a class of organizations,
whether already documented in the literature or to be documented by future
studies.
However, the biggest difference lies in social scientists' governing ethic.
Dingwall (1992) proposed that this could be described as 'fair dealing', the
notion that the researcher's role is not to sit in judgement but to represent
as dispassionately as possible the contribution of each participant to the
production of the setting that is being studied. The resulting analysis may be
a source for moral outrage but it should not be a vehicle for this: effective
reform demands an understanding of how morally outrageous things come
to happen, which is rarely the result of deliberate wickedness at all levels. We
see a concern for this aspect of fair dealing in virtually all of the qualitative
studies discussed in this chapter.
This chapter has added another dimension to Dingwall's (1992) initial
conceptualization of qualitative research as fair dealing. We have called the
additional dimension 'best practices' to highlight qualitative researchers'
distinctive vantage point for observing how solutions to organizational prob­
lems are often already evident in the everyday practices of organization
members. Ker Muir's (1977) discussion of the distinctive orientations and
practices of professional police officers is an example of how qualitative
researchers can identify best practices hidden by dominant problematic
patterns, effective work within a troubled environment. Orr's (1996) study
of copy machine technicians, on the other hand, shows how qualitative
research can reveal widespread (but officially unrecognized and discouraged)
best practices developed by organization members in order to achieve
organizational goals.
Fair dealing in qualitative organizational research involves dispassionately
analysing how organization members contribute to the production of efficient,
effective, equitable and humane work processes and relationships as well as
338

339
 4 )£11&

GALE MILLER, ROBERT DINGWALL AND ELIZABETH MURPHY " THE WIDER COMMUNITY i

U SING QUA LITATI VE DAT

A AND ANA LYSIS
R E FE RENCES Mi shl er, Ell iot G. ( 979)
, ' . ng 111
'Meam .
context.· I s there any othe
Educational Review, 49: 1-1
Bosk, Charles 1. (1979) Forgive and Remember: Managing Medical Failure. Chicago:
9. r kind ?', Harvard
Orr, Julian E. (19 96) Talkin
g Ab out Machines.· An Eth
University of Chicago Press. NY : Cornell University Pre nography of a Modern Job.
ss. Ithaca,
)
I attem, Michael Qumn ' (2002) Q,Uall.tatlve
Research & Evaluation Met
Brown, John Seely and Paul Duguid (2000) The Social Life ofInformation. Boston: Harvard .
Business School Press. CA: Sage. hods . Beverly Hills,
Chawla, Sarita and John Renesch (1995) Learning Organizations: Developing Cultures for Rawls, An�1 (1987) 'The lnt
eraction Order Sui Gener · .
Tomorrow's Workplace. Portland, OR: Productivity Press. Theory , Soc1010gical Theo IS." (,oH.man" s
ry 5: 136-49 Contnb utlOn to S ocial
.

Dalton, Melville (1959) Men Who Manage: Fusions ofFeeling and Theory in Administration. Roy, Donald F. (1959) "'B
New York: John Wiley & Sons.
a�ana Time';.' J ob Sahsf . .
Human Organization, ] 8: achon and Informal Intera
158-68. ction ',
Dingwall, Robert (1992) 'Don't mind him - he's from Barcelona: qualitative methods Shaw, Ian (19 99) Qualitativ .

e Evaluation. London .. Sage . ..

in health studies', pp. 161-75 in Jeanne Daly, Ian McDonald and Evan Willis (eds), Shaw, Ian and Joyce Lishm
an (ed s ) (1999) Eval, uation
Researching Health Care. London: Routledge. Sage. and Social Work Practice.
London:
Dingwall, Robert, John Eekelaar and Topsy Murray (1983) The Protection of Children: Sta cey, Ralph D., Douglas
Griffin and Patricia Shaw (20 .
Systems Thinf,(zng' .? . London : Rout
State Intervention and Family Life. Oxford: Basil Blackwell. Fad or Radical Challenge to g0) Complexlfy and Management:
A
Stake, Robert E (]. 995) The
1"110usan d Oaks,
Dingwall, Robert, Elizabeth Murphy, Pamela Watson, David Greatbatch and Susan '
c
ledge .
J . (20 01) The' Paradox. of Con
. rt of Case Study Research .
trol in 0 a ."zatlO s .Lond
Parker (1998) 'Catching Goldfish: Quality in Qualitative Research', Journal of Health Streatfield, Philip'
:f. ;
CA: Sage .
Services Research and Policy, 3: 167-72. Vaughan, Diane (1983) ' on: Routledge.
Con trolling Unla wful Behav l . n.
Fisch, Richard, John H. Weakland and Lynn Segal (1 982) The Tactics of Change: Doing Mi sco ndu ct. Ch ica go : Un i . oct al Struct ure and Corpo
rate
Therapy Briefly. San Francisco: Jossey-Bass. -- (1 996) The Ch . iversity. ()f Chlca
allenger Lau nch Decisions
'
' go Press.
. : Risk Technology, Cult..
ressY
Glaser, Barney G. and Anselm L. Strauss (1967) The Discovery of Grounded Theory: at NASA . Chica go: University of Cl1l. cago P .
ure, and DeVia nce
Strategies for Qualitative Research. Chicago: Aldine. Wenger, Eti'enne (1998) Com
mu nit ies of Prach.ce.: Lear
nzng, Meam.ng, and lden
iversity press.
Hammer, Michael and James Champy (1 993) Reengineering the Corporation: A Manifesto Cambridge: Cambridge Un .
tity.
for Business Revolution. New York: HarperBusiness. .
We�ger, Ehenne, Richard
McDermott and Willia m
Kanter, Rosabeth Moss (1977) Men and Women of the Corporation. New York: Basic Books. Commumtzes of Practice: A M. Snyder (2002) Cultivati
Gu ide to Managzn ' g Kn. owledge. Boston : ng
Kaplan, Bonnie (1986) 'Impact of a Clinical Laboratory Computer System: Users' School Press. Harvard Business
Perceptions', pp. 1057-61 in B. Blum and M. Jorgensen (eds), Medinfo 86: F1fth World Wiener, Carolyn (2000) Th
. Hospitals. Hawthorne,
e Elu siv e Quest.. Accounta .
Congress on Medical Informatics. Amsterdam: North-Holland. Aldine de Gruyter. bllzty zn '

N Y:
-- (1987) 'Initial Impact of a Clinical Laboratory Computer System: Themes Common
to Expectations and Actualities', Journal ofMedical Systems, 1 1 : 37-47.
Kaplan, Bonnie and Dennis Duchon (1988) 'Combining Qualitative and Quantitative
Methods in Information Systems Research: A Case Study', MIS Quarterly, 12:
571-86.
-- (1989) 'A Job Orientation Model of Impact on Work Seven Months Post
Implementation', pp. 1051-5 in B. Barber, D. Cao, D. Quin and G. Wagner (eds),
Medinfo 89: Sixth World Congress on Medical Informatics. Amsterdam: North- Holland.
Ker Muir, Jr, William (1977) Police: Streetcorner Politicians. Chicago: University of
Chicago Press.
Kunda, Gideon (1992) Engineering Culture: Control and Commitment in a High-Tech
Corporation. Philadelphia: Temple University Press.
Lave, Jean and Etienne Wenger (1991) Situated Learning: Legitimate Peripheral
Participation. Cambridge: Cambridge University Press.
Lipsky, Michael (1980) Street-Level Bureaucracy: Dilemmas of the Individual in Public
Services. New York: Russell Sage Foundation.
Marlaire, Courtney 1. (1990) 'On Questions, Communication, and Bias: Educational
Testing as "Invisible" Collaboration', pp. 233-60 in Gale Miller and James A
Holstein (eds), Perspectives on Social Problems: A Research Annual, Vol. 2. Greenwich,
CT: JAI Press.
Miller, Gale and Robert Dingwall (eds) (1997) Context and Method in Qualitative Research.
London: Sage.

340

341
 WHO CARES ABOUT 'EXPERIENCE'?

because they have to rely on more remote, inferential empirical methods and
materials. (Denzin and Lincoln, 2000: 1 0)

Denzin and Lincoln's portrayal of what qualitative researchers 'think they can
[do]' (above quotation) is a deadly accurate characterization of much contem­
porary qualitative research. It appears to differentiate us beautifully from
those benighted number-crunchers whose concern for mere 'facts' precludes
a proper understanding of what Denzin and Lincoln call 'the actor's perspec­
tive'. By contrast, our research concerns itself with '[the] perspectives of
the participants and their diversity' (Flick, 1998: 27) and attempts 'to document
ab o ut the world from the point of view of the people studied' (Hammersley, 1989:
165). When we study entities like organizations, then, the key question is: 'How
qu alitative to get at and document the lived experience of organizational members?'
Mi ssing i ss ues in
research*
(Eberle and Maeder, 2002).
This attention to the 'perspective', 'point of view' and 'lived experience' of
David Silv erman
the people we study reveals a rarely challenged consensus about the nature
of our enterprise and its analytical targets. Although, as Flick points out (1998:
17), it is often associated with the symbolic interactionist tradition, it extends
itself much more generally throughout qualitative research.
For instance, within most organizational ethnography, 'experience' is the
much. Our prized object. As Aldrich suggests:
m uch abo ut peoP Ie i n these days; we hear too ry James,
We know too . onalities. (Hen
ouths are stuffed with pers the various interpretive views have in common their focus on an actor' s perspective
ears , our minds, our m
18) on life in organizations. (Aldrich, 1992: 23)
The Portrait of a Lady, 2

� ��� �;
. d m e Merle as an isolated
figure;
Yet what Aldrich calls 'the actor's perspective' is a very slippery notion as his
Isabel found it
onl y m
iffi
her
cult to
relatlO
thm
. ns : spo
e
f Uow-mortals. . . . Madam none
ke
Merle
intellectual ancestor, Max Weber (1949), was well aware. For instance, it is by
�
she existed her nature
she. Sll wa deep- and
was not superficial - not aus �1t s� ke a c�nventional language. 'What no means clear how such 'perspectives' relate to our actions. Indeed, is there
the less in her behaviour
bec � .
sald I sabel . 'She has the
good. taste a 'point of view' or 'perspective' lying behind every act?
is langu age at all b ut a co nv entlO n?
e peopI e I have met to
express herself by on
, aI
gm Most qualitative researchers who champion the subject's point of view or
not to pretend, like som privilege experience simply do not question where the subject's 'viewpoint'
signs. (ibid. : 167)
are comes from or how 'experience' gets defined the way it does by those very
alita tive re search . GlV
.
en
? the dispa r ate nature of the ditopngic,tehoxtbwook individuals whose experience we seek to document. Do these not emerge, in
What is qu r such a quesflOn.';> Rather than add to the unenappears to be a some way or other, from the varied contexts out of which we 'draw from
we to answe chapter will focus on one feature where there organizations, experience' to convey accounts of who and what we are?
debates, this sm. g examples drawn from ethnographic work on r a reasoned A telling example is provided in an anecdote about one of Jay Gubrium's
consensus. Ushow why I dlsa . this consensusveyandSacoffe doctoral students (Gubrium and Holstein, 2002: 21-2). The student inter­
I will then drawing upon thegrenege withl e �t ed wor k of Har ks (Sacks, 1992; viewed pharmacists who had engaged in substance abuse. His aim was
alternative 1998; Bake�, �hi. s volu�: � wor�. T�ee to understand how those who 'should know better' accounted for what had
Silverman,sensual posItIon IS. wel e out in. a recent authoritative happened to them. As it turned out, what these pharmacists said closely
The con oductl. On to the S econd E d'Itlon f The Handbo 0 ok of Qua lIta tIv fitted the familiar recovery rubric of self-help groups. Indeed, many had
Intr
Editorial ts it this way: attended groups like Alcoholics Anonymous (AA) and Narcotics Anonymous
Research pu . d'IVI' duaYs (NA). So in what sense were these accounts the pharmacists' 'own' stories?
.
anhtat.l' ve . researchers are
concerned with the m As Gubrium pointed out, do these stories not 'belong' less to individuals
Both qual itative and qu . tatIve mvestigators think they can get closer to at
the
than to particular organizational discourses which are merely 'voiced' here?
�
oint of view. Ho wev er,
ctor' s perspective throu
qualI
��
gh deta d int I �: �:�:�
. . and observation.
ture
They arg ue
their subjects' perspec
th
tives By contrast, a concentration on 'lived experience' (rather than 'voices') can
are se om a I
quantitative researchers
343
DAVID SILVERMAN POSTSCRIPT "

WH O CARES ABO UT 'EXP ERIE

N C E '?
lead to an essentialist, 'romantic' conception of inner meaning (Silverman, You might say: how on ear
1989). We can see this romanticism at work in Aldrich's references to: by her son's release from inttherncan a mother be anyth'Ing other than rel
1 . ieved
if such a mother wanted her .s on to s�t?. S.ur� y, It would only be newsworthy
m e
the expressive side of participation (which) is as important as the task-related side, release? ay In amp X-Ray and was upset by
his
as organizations do not operate solely on the basis of a rational economic model. Yet our me dia (an d, to be fair, most of , h S l1p. pets of
(Aldrich, 1992: 25, my emphasis) rently 'lived experience' . The fact tha us) d em' and suc
t the t�:n out to be entIrely predicapp
� a-
have the appropriate response . Th�Init1 is Je . nded of us all that we should
seems to be of no account . Indeed tab le
We might ask Aldrich: why are 'task-related' activities necessarily moulded
according to an economic model; why do we have to assume that activities so 0ften tell Int
· erviewers that 'it has notforsunInsl tan
<., ce' how suc cess fu1 athI etes
are guided either by a 'rational' logic or a logic of sentiment? Does this not be seen like we p erceive ourselves, ' '
In y et, . After all, they need to
return us to the 1930s and the Hawthorne studies?l fortune . as d ecent people, surprised by go od
<.

Two examples of organizational ethnographies will serve to illustrate my Rarely do es the absurdity
point.2 Hassard (1 993) calls his study of a fire station an 'interpretive ethnog­ after the rescue of some menoffrothe se kind 0f r�spo�ses SIn . For
' � In. instance,
raphy' . His data 'involved accompanying firefighters during the working interviewer asked a mine official: rr a1rec ent mme dLSaster In the USA, a TV
day and asking them to explain their activities before, during and after each
event' (1993: 99) . Hassard claims that this represents an 'ethnomethodological' Int: How did peo ple fee l afte
stance (98) which: Offic ial: Act uall y I was too
r the me n we re rele ase d?
bus y talk ing to peo ple like
you .
portrays the everyday work of a fire station in terms of how firefighters make sense (US TV report)
of and enact the task system. (101) This is a rare example of an interview
The mm . e officia ee refusing to PlaY the Ex ene. nce Gam
l shows us that this kind 0f feeh . ngs dIscou � e.
Yet his use of interviews, analysed in terms of 'recurrent theme(s)' (100), hardly media demands . rse IS a product of
challenges the consensus I have outlined since it fully concurs with the latter's
focus on how people 'view' or 'experience' the social world .
Occasions like mine accidents and mu
marvellous grist to the media mi ll Re fe� . rgdto
nn
�
ers partIcu. l rly
� of children) are
Linstead and Thomas's (2002) use of a 'poststructuralist feminist read­ aftermath to the murder in Soha�, E t e reporhn� of the immediate
ing' to define their research on middle managers might seem to threaten the ng and, of two young girlS, a critic wr
ote:
consensus more effectively. Yet, as it turns out, like Hassard, in this study
interviews are used and analysed in terms of themes.
Ho w do you feel about the abd
Come on, how do you really fee
l? .
�
uction and murd e of those two
ten-year-old girls?
In both my examples, apparently non-orthodox positions happily fall and insensitive question . So wh ' Yes, you 're ng t, It's a crassly stupid, pointless
y have mueh of· the media spen
within the consensus. Although ethnomethodology, poststructuralism and demanding answers to that and t the 'past fortnig
feminism would appear to have little in common with qualitative research's beyond? (Hume, 2002) ' ns from everybody 111 Soham ht
similar questIo
and
symbolic interactionist roots, here they do, sanctified by the siren call of 'lived
experience' . 3 The roots of such a stubborn consensus surely extend beyond Mick Hume's questions should e �
social science theories and say something about the world around us . One uallY be po sed to researchers who
than a century ago , in the wo rds of �rh?S, as Henry James suggested more
tempted by the lure of 'experien ce' are
inescapable fact about the cultural milieu of the Western world is that 'lived
too much about people in these dah� c aracter Ralph T.ouc�ett, 'we know
experience' must be pursued and interrogated. One aspect of this is seen in
the work of the 'psy' professions, notably in the methods of the skilled start of this chapter . ) Put at its star � ?'
?e
�� e full �u�tatlon IS given at the
counsellor (see PerakyHi, 1995). 'experience' and 'feelings' w1len, In st st
. so ' . �uahtatlve research privilege
Of course, not everybody visits a counsellor or psychiatrist. But many of .lm atives as a soap opera or a therapydo mg. ' It is respond'Ing to the same
us avidly watch 'reality' television shows, soaps and chat shows in which .per .
fhere IS, of course, an obvious resp seS SIOn?
people's 'experience' is very much the target (see Atkinson and Silverman, ons e t0 my rheton. cal questIOn .
1997) . TV news programmes also increasingly cater to our demand to get
necessarily. Let us take one exam Ie
dispenses with interviews and mKes of ;� �;;. a � .f�h. �
I �
: no t
al. eth�ography which
inside people's heads. Take the example below, garnered more or less at and expressive organization. n IC 1- 1 e Istmctlon between task
random: Zimmerman (1992) was concerned . nal
the talk occurring in calls to WI'th tlle mt
' eractIO organization of
Mother 'relieved' at son's release from Camp X-Ray. (BBC TV teletext headline)
US
by the emergency services . His 1 e�ergency telephone numbers operated
91
researc was based on data from three cen
tres
344

345
 WHO CARES ABOUT 'EXPERIENCE'?

DAV ID SILV ER MAN

POS TSCRIPT

sed on calls to unlike Aldrich (and the Hawthorne researchers) who place human beings and
. West, two on the W est Coast).. His stu dy focu (CAD). machines in two different boxes, Zimmerman shows us how machines corne
(one in the MId-
ntre s w h '
IC h h ave a com puter-assisted dispatch talkem
syst
in terms of· to have a presence in human interaction (see also Heath's chapter in this
two of the ce
(CTs) ha � e t��tt. e d to callers' (Cs) volume and Heath and Luff, 2000).
At all tim es, cal l-tak ers
CA D co eS ' : eans that res ponse toke ns from
Of course, Zimmerman's study did not emerge out of a clear blue sky.
und �:�he computer keybout to say. Hence
what is re por tabl e via the boa rd. CTs must
are ofte n re plac ed b y th e s ? Several decades ago, the pioneering work of Harvey Sacks set out the template
the CT an to what they are a C's talk sug ests for a new model of qualitative research. I will now, therefore, provide a taste
also orient both :o.,:h� t CsenareU�;!e�� l r their monitoring of the � of what Sacks has to offer (for a book-length introduction, see Silverman, 1998).
at mg mds
they m ay defe r mitl
very of a
C)
pIec e
.
of pertment m ' f�rmation. Som e of thes e k
Throughout, I will attempt to show you that Sacks is more tha11 the founder
are onented t0 bY s .
the up com ing deli
of organizationa1 , so1utiorns'teleph ne ca11s, the opening sequence d oes ypl- n�t of what you may take to be a sect ('just one of those ethnos') but has much to
teach researchers who will never want to do conversation analysis or
So, here, unli. ke othe g or a �how are you7'. sequence. Both parties t ethnomethodology. In this spirit, let us look at Sacks's insistence that 'activities
routinely contam a gre,etin re so � f0 r the call' sequence as soo n as pOSSI'bI �. are observable' and contrast it with ethnographers W110 privilege 'experience'.
cally see k to rea ch the � ract 1 b eI ow , where a caller is reportmg an
We see thes e feature s m Ext
incident: ACTIVITIES ARE OBSERVABLE

� l@Q
used)
d f keyboard bein g
I thin k it's uh (.)
. '
As I have already noted, a popular activity in everyday life is to wonder about
�: �� ���� �
=
er 's uh (0.2) oh
��� people's motives. Indeed, in the case of talk shows, the motives of the rich,
------
.
[ pull ed u p here uh In fr
(.) hh jeepster th a t - -
(kb - ,
guy s th a got out, I
- --- famous or just plain unlucky or deviant become a central topic. Yet, in many
respects, social science has picked up this habit, taking as its task the revelation
: ""An the re's abo ut (0. 1 )1 five Nih gro
C of other people's 'motives' and 'experiences'.
- --- --- 1
��� � ��� �
r t l (kin about (0. 1 ) hh � �: �.)_ �_t����
g i t next
Even in the 1960s, Sacks seemed fully aware of these issues. His kin d
(0. 2) [hh a1n I JUs
- - - -
[- - - - - - - - -- . , of social science always turned away from the insides of people's heads and
ding uh fur uh flgb1 towards their observable activities. Like Madame Merle, as viewed by Henry
apartm ent buil [- - - 1
James's Isobel Archer (see quotation at the start of this chapter), Sacks saw
ch eck it out that we exist in our relations with others and express ourselves through a
am goi ng dow n tuh (Zimm erman , 1 99 2: 4)
conventional language. In this sense, Sacks was a self-proclaimed behaviourist
who announced that his task was to elucidate how people did whatever they
extract:
Zimmerman notes about this l the call-
did. As he put it:
. n of C ' s count of the trouble . . . (on line 8) CT use of
� For Members, activities are observables. They see activities. They see persons doing
�
At the pos sible con cluS I ard activity. Aft er a pa .
say s noth ing whIle she co tinu. es her keybo abs enc e of a receIpt intimacy, they see persons lying, etc . . . . And that poses for us the task of being
taker] � . ' ' hich in th e
�
be onente to the possi
iates (an) elabo ratlO pro blem behaviourists in this sense: finding how it is that people can produce sets of actions
0.2 second, C init bility of some
m ent fro m C T, may that provide that others can see such things. (Sacks, 1992, Vol. 1 : 1 1 9)
of acknowledg
427)
with his narrative. (1992:
vities As examples of such 'sets of actions', Sacks offers 'describing' and 'question­
res pon se tok ens, one ex F ected feature of hearers' acti ordinary ing'. These are interesting activities because each may be seen as a resou rce
This absence . of . t
IbIe tu. rn-transition poinZIm in
.
a tu rn a t any giv en P OSS m er man for social scientists, e.g. ethnographers 'describe' cultures and 'question'
when not tak mg
obs erva ble con sequ � n�:� these calls. As as indicating informants. However, Sacks wants to make both activities a topic by examin­
conversation , h as
such � response �ay be h eard byade
C
s. These a:.e ing them as forms of behaviour which, through some methods awaiting
suggests, the absence ofThu � t�h ed and do wngr
ng that It
port. s It can le d g es inspection, are produced and recognized.
a trouble in the re 1lOWn) , when C hed
ges, stati
(da ta no S It follows that how societal members (including social researchers) 'see'
seen Iess tha n ten line s later
particular activities is, for Sacks, the central research question. In this respect,
might not be any, thing'.oach to the use 0f a computer-assiste. d tec h nology
. rvI' ews, together with Garfinkel (1967), he offers a unique perspective in social science
Zimmerman s appr -structur:�s
. ted Without usmg mte which 'seeks to describe methods persons use in doing social life' (Sacks,
reveals how a par ticular tas � �;;� m � uni cate and m ake dec isions
me h 1984: 21).
he is able to show us inHen re�l tl
tag e an � p� tt !r in this volume). Put another way,
(see the chapters by 347

346
i.
WHO CARES ABOUT 'EXPERIENCE'?

DAVID SILV ERMAN 0 POSTSCRIPT

you offer a proper warrant for askin O '· 4-5). By contrast, providing a slot
cribe' a nd ' qu es tio n', very often they are tacitly usinialg for a name cannot be made accountabYe · So t. o .answer a phone with your name
When researchers 'des are to study such methods, it is, therefore, cruc has a function in institutions where 0bt ammg callers' names is important
members' methods. If we gra nte d what it is we ap pear to b e ' seeing' . As Sacks (1 : 5-6).
that we do not take for Of course, the fact that somethin m ay properly l:appen once a slot has
says:
the been created does not mean that it wilfha ppen. Take thIS further example cited
p reparatory to solving by Sacks:
ing up w hat it i s th at seems to have happened, ceiv abl y h appen decid e
In sett ur notion of what could con
happened. (1 15)
not let yo
[research] problem, do
3
for you what must have
A:
ld conceivably happen'
This is M r Smith may I help you
tha t our 'no tio n of wha t cou B: I can 't hear you
Here Sacks is telling us m our u nexamined members' knowled ge. Instead, A: This is Mr Smith
is likely to be drawn fro members use -

d to proc eed m ore ca utio usly by examining the methowds hat Sacks meant
B: Smith. --

we nee
duce ac tiviti es as obse rva ble and reportable. Let us see (1 : 3)
to pro of an organization.
by looking at his studynscr
ery fi rst tra ibe d le cture, given in Fall 1964, Sa cks begins with
lected at an
�
Sacks's two procedural rules do not mean tha s�eake:s are automatons. What
seems to happen in Extract 3 is. that B�s reP Y I can t hear you' means that
In his v
tion on telephone conversatio ,col
ns
data fro m his PhD diss erta
hos pital. In the follo wing two extra cts A is a member the slot .for the other party t gIve
° thelr' name i " mlssed . 'T'Jus. does not mean
s

emergency psychiatric ut themselves

.

at the hosp ita l an d B can be somebody either calling abo that thelf name is 'absent' but rather that the place where it might go is closed.
of staff else. As Sacks puts it:
or calling about somebody
It is not simply that tlle ca11er Ignores
' what they properly ought to do, but something
1
A: Hello :? ���;
the return name fits is ne;er ;pe d. l
rather more exquisite. That is th h ays of providing that the place wh ere

B : Hello (Sacks, 1 : 3)
Put at its simplest, researchers must be ver. Y careful how they use categories .
For
. cinstance, Sacks quotes from two l'mgUlsts who appear to have no problem
may I hel p you m haractenzing
' particular (in nted) uttera� ces as 'simple', 'complex',
2
A'. This is Mr S mith
B: Yes, this is Mr
Brow n (Sacks, 1 : 3)
'casual' or 'ceremonial'. For Sa:� s, such rapId characterizations of data
assume
ons rst B seems
ab out these extracts. FiSo,
es tw o init ial obse rvati in Extract that we can know that WI'thout an analysis of what it is (they) are d omg.
' (1: 429)
Sacks mak to th e fo rm at pro vided by A's first turn. of nam
to tailor his u tte ra nce es . Or,
'h ellos ' and, in Extract 2, an exchang'ewhere Such an analysis needs to locate partiCU1ar utterances in sequences of talk (1:
L we get an exc hang e of
that there is a 'procedural rule 430 and 622). So an ethnograph er wh0 reports that (s)1le 1leard someone tell
as Sacks puts it, we might say
can choose their form of
a 'story' only raises a further ques t'IOn: how the eth�ographer (and presumably
.
the other uses. (1 : 4)
who spe aks first in a telephone conversation the members of the group studied) heard an actIvIty as a 'story'.
a person
ress , and . . . the reb y choose the form of address
add
ea hange occurs
ch part or turn of the exccalle
eco nd obse rvati on is that d a 'unit' C O M M ENTATOR MAC H I N E
Sacks's s
a pair (e.g . Hell o-Hel lo). Each pair of turns may bend sets up an
as part of onstitutes a 'slot' for th e second
a
on, Through Sacks, let us now look more closely at. h ?w ethnographers use
in which the first turnt thcis slot may properly contain. Given this expectativing mformants to support their findings' fo� mstance ' m mterview studies or by
expectation about wha act B's name (as in Extract 2) without ever h a ids means of 'respondent validation'. B d�mg, so, Sacks suggests that they treat
A is usually able to extr beauty of this, as Sa cks points out, is that it avome­ people's ability to give and receI'v'".... J
senSI.ble descr'IpfIOns m
. d'Ifferent contexts
to ask for it directly. Thequestion might create. For instance, if you ask sothat .
as entIrely non-remarkable . As he puts It:
a problem that a direcmay t uire
r nam e, they pro perly ask 'Why?' and, in this way, req
one thei 349

348
 W H O CARES ABOUT 'EXPERIENCE'?
POSTSCRIPT
DAVI D SILV E R MAN
. , . t persons produce Moreover, such neglect cannot be remedied, as some researchers claim, by
not to clarify these, o� t
' a r is' even If It can b e Sald tha assembling panels of judges to see if they see the same thing (e.g. inter-coder
The essential 'me ssag e'. of t h
or
� r:
Is
l
'
e task of sociology is
7�
descriptions of the s OCIa l w :. . ribe them . (Sacks, 1963. agreement as a basis for claiming that one's descriptions of data are reliable).
on the reco rd ' or to cnhClze them, but to desc Such agreement offers no solution because it simply raises further questions
'get th em
. tly famous for the example 0f a machm ' e with. its about the ability of members of society to see things in common - presumably
. rIgh by using the 'etcetera' principle as a tacit resource (see Clavarino et al., 1995).
Sacks's (1963) pap er IS l machine mIght
.Ilt-i' n commentator . He te1ls us that this hypothetica Sacks's problem is how we can build a social science that does better than
own bU m an . the following terms:
m this. In some way, we must free ourselves from the 'common-sense perspec­
be des crib ed by the lay
. g s, ome J. Ob , and the other part tive' (10--11) employed in our use of 'undescribed categories'. For Sacks, the
. engaged in dom solution is to view such categories 'as features of social life which sociology
one par t IS For the commonsense
I t h as tWO parts;
ra es � h '
at th e
.
£lrst part d oes
r machine', its parts th
e , must treat as subject matter' rather than 'as sociological resources' (16).
b e ca11ed a 'commentato
. · ·

nar
.

synchronically �
rrllg ht
perspective the mac hm e What looks like a complicated theoretical solution turns out, however,
' parts. (5) to involve a quite straightforward direction for research. We must give
doing' and 'the saying
in� is to up defining social phenomena at the outset (like Durkheim) or through
. g ethnogr�pher the 'saying' part of the m ach mg of the accounts that subjects give of their behaviour (Sacks's 'commentator
akm
For a native-spe od, poor °r u�nlca . ' 1 description of the actual work
ation machine'). Instead, we must simply focus on what people do. As Sacks
a go explan
b e analysed as
(5-6 ). H ow ever , S ac (S pm 'nts out this sociolo gical puts it:
. machine nowledge..
�� :des off two kinds of unexplicated k
it emits
whatever humans do can be examined to discover some way they do it, and that
. ' common with the machine the language way would be describable. (1992, 1 : 484)
(a) k nowmg In
and d oing . (6) We now see why Sacks should have referred to sociology's 'peculiar stance'.
owing in some language wh at the machine is
(b) kn Quite properly, he suggests, sociology seeks to be a science. In doing so,
achi . ne IS. do . n�, ultimately depends upon a set ofd it needs to seek a 'literal' description of its subject matter. However, this
at the m
But to know 'wh onsense a;:um�tl�� s b ased on everyday languagecrIb a� search is undercut, in Sacks's view, by sociology's use of concepts ('a descrip­
pre-scientific, com m to 'des e
'fac ts' fro m ancy . t follows that our ability shou ld
tive apparatus') based on unexplicated assumptions (1963: 2).
employed to sort lay people or SOCIO . 1ogI'sts, 'is a happening' wh ich
. 0 By contrast, Sacks asks us to make this descriptive apparatus our topic. For:
social life' , w1leth er as SOCIO . 1 og ' (7) to describe rather than tacIt1y t use . Te
properly be t�e 'job ofon the PrIO y of describing the everyday hIS '�roce�� whatever humans do can be examined to discover some way they do it, and that
ste nce ritY posItIOn
Sacks's insI
ass em bling cases of t h e c1ass' radically separates way would be describable. (1992, 1 : 484)
employed for m romanbc ethnographers. � for
b oth from Durkheimarandntfr�lH ren both Durkheim and Goffma� �ak Sacks concedes that this kind of research can seem to be 'enormously laborious'
Despite the ir a p ? � �l : t ��ich eo Ie respond (e.g. 'SU lc l ) or
de (1: 65). However, he denies critics' claims that it is trivial. You only need to
granted som e soc Ial
l
e
a
��
; rI ?) identifi!d 0; the basis of tacit
e
com m onse�s�
It
look at the ability of both lay persons and conventional researchers consistently
describ.e a pro cess
eir
(e.g.
com m on
�
faIl
?
mg IS, as Sacks puts it, that they rk W
wo to find recognizable meaning in situations to realize that social order is to be
reasomng. Th . , found in even the tiniest activity. The accomplishment of this 'order at all
'undescribed categOrIes : points' (1: 484) thus constitutes the exciting new topic for social research.
as ap,rear Beginning with the observability of 'order at all points', our first task should
. bed categ ry is to write descriptions such
To emplo y an unde scn � . re are pictures of objects. be to inspect the
wIth sen es 0f word s the
rsp erse d
in children's b ooks. Inte
. tS get by throu h simply 'pointing' at familiator
ogIS
collections of social objects - like 'How are you feeling?' - which persons assemble
ciol able
so e? definition) " So theybare
For Sacks, m ost . to do their activities. And how they assemble those activities is describable with
at p h '
I l OS O P her s call 'ostensIv ng y ' nvo king respect to any one of them they happen to do. (1: 27)
objects (wh Sa; ks" s com men tato r m achi ne' is 'doi :
give an account of whatws abo u t how thi ngs are in society usmg wheatr
'what everybody k noto as the 'etcetera r. . l ' They thus pretend to so off This version of our research task clearly sets Sacks apart from most quali­
Garfinkel 7 (196 ) refe rs
nom ena w : IC ���� e � ls the ir 'negl ect ( of] me tative researchers. However, despite his radical perspective, Sacks was too
'literal' description of phe (13).
sophisticated a thinker to believe that any approach, including his own, could
�ndetermined set of their features' 351

350
DAVID SILVERMAN "' POSTSCRIPT
WHO CARE S AB OUT 'EXP
. predecessors . I et us now consider what Sacks believed ERIE NCE'?
learn nothing from Its To solve this problem, lice 'lea
we can still learn from more convent�1onal forms of ethnography. appearances' (Sacks, 1972:po284 rn to treat their beat as a territory
of normal
appearances as 'incongruities' ). No w they can treat slight variations in
normal
assumption of the appearances worthy of investigation, working with the
T H E VALU E (AN D L I M ITS) OF ETH N O G RAPHY
Sacks's point of departure froofm'normal' crimes (d. Sudnow, 1965).
work on ceremonial orders. Al Goffman can be seen if we compare their
Sacks's lectures make. n:any fa�ourable references to the tradition of ethno­ directly from Goffman, it soon bethough parts of Sacks's early lec tures draw
graphic work . that ongmated n the 1930s in the Sociology Department of 'ceremony' not by reference to comes clear that Sacks wants to understand
the UniverSIty 0f Ch'Icag0 . ThIS. wor1( represented the first flowering of an 'frames' but to the sequential anco ncepts like 'impression mana gement' or
empirical sch001 0f soc iology
. theconcerne d WI' th observing what it terme d the volume). alysis of conversations (see He
ritage, this
'subcultures' to be found m bUl'ld'mgs and the streets of the modern CI. ty For instance, we know that the
(see Hammersley, 1989). . 1 tack, Sacks found much to admire in 'fine'. This means that if you want proper return to 'how are you feeling?' is
D . te taking a different theoretlca you have to request permission to treat it as a question about your feelings
the ��r;ago School's attention to detail. As he comments. next party may say 'that's alrigh(e. g. by saying 'it' s <t long story' where the
c

. aSIde. the o d er ethnogra hic work in sociology, I would treat it

ha s to lie' because people attend t, I have tim e') . This means that 'everyone
Instead of pushmg and, in part, produce answers by to 'the procedural location of their answers'
� ��� f�f�
. . �lzmg
. .
l S .
.
. where criticizing is glvmg
as the only work worth cnh C some may have for next actions that mareference to 'the various uses that the answer
. . something.
.
) r ance of the works of the Chicago y be done' (Sacks, 1992, Vol. 1: 565
dignity to So, for e�amp C, This focus on 'procedural organiza
soclOloglSts IS t11at theY do contam a lot of m
' formaHon about this and that. And this-
to separate the 'ritual' and 'syste tion' means that Coffman's (I 981 ) attempt
).
.
and-that is what the world IS made up 0f. (1 992, 1 : 27)
been a non-starter for Sacks. Co m' requirements of interaction would have
Sacks has the focus of a 'systems ntrary to Goffm an's (1981) suggestion that
Like the older . ethnogra�hers, Sacks re'ected the crass empiricism of cer.tain rule-governed but rule-guided. engineer', Sacks shows that behaviour is not
kinds of quanhtat'Ive so. clOlogy . In partlcu) 1ar, as we have seen, its assumptlons you will be held accountable for In this sense, you can do what you like but
that research IS· based on. finding some m
_.

. d'Ices and explaining why they nse . the implications of your actions.
unlike an engineering model, soc
and fall by ex post facto mter�r�t�tio.ns of si nificant correlations.
Moreover,
interaction, an 'offshoot of a machi ial ord er is merely a by-pro du ct of social
Sacks had taken courses. wIt. r�mg Goffman at Berkeley in the early 1960s in particular' (Sacks, 1992, Vol. ne designed to do something else or nothing
and found. much t0 admue m .hIS wor1(. UnI1'ke Bales , who mainly used Unlike Goffman, Sa ck s's sto2:ck24 0).
laboratory data, Goffman was mteres ' the complexities of natura 11y
te d m examples but detailed transcriptio in trade was no t anecdotes or invented
occurring behaviour. Unlike Homans, 1le did not want to 'correct, everyday this kind of data, Sacks claimed ns of tape-recorded conversa tions. By using
understand'mgs b t to show the compIex way in which they functioned. much more detailed way. that he was able to look at intera
ction in a
In fact, Goffman U.CItes as. hIS" �ntellectual ancestor the German SoClO . Iog . st
� A good example of this is When a
Georg Simmel, commendmg hIS penetrating analyses of the 'forms' of socIal that Sacks's example of a visito student at one of Sacks's lectures suggests
interaction. fully explained by Goffman's con r interp reting appearances in a
. household is
So Goffman's work was Impor ta�t to Sacks ' Like Goffman, Sacks took
. usly referrmg people 'give off'. Sacks replies cept of the incongruous impressions that
Simmel'� w,ork very seno t0 h'1m as 'one of the greatest of all Goffman 's approach. As he putsthat, in fact, he has a basic difference from
sociologIsts (Sacks, 1992, VoL '2 .' . 132). Like
c

Goffman, Sacks had no interest m . it:

building data- f�ee gr�nd theones or m . resea rch methods, like laboratory Goffman tal ks abo ut respo
nses to incongrUity but he
I think I'm beginn
studies or even mtervlews, which abstracted people from everyday cor:texts. incongrUity is. That's what does no t tell us what
Above all, both men marvelled at the everyday skills through which partlcular Ho w it is that one sees it. He ing to see here in this stuff
ha s no t analyzed how it is tha [datal .
appearances are maintained. , what makes it an incongrui t you do 'an incongruity',
ty. (Sacks, 1992, VoL 1: 92)
We can catch sight of Sacks s use of G0ffman's ideas in his paper 'Notes on Another way of putting thi
Police Assessment of Moral Characte,r, (Sacks' 1972) which was originally of social order in far greatesrisdetha t Sacks wanted to study the local
written as a course p aper for Goff-man s cour se at Berkeley in the early 1960s tai production
(Sacks, 1972: 280n.). For Sacks, po l'Ice fficers face the same kind 0f problem serious work paid attention to det l than Goffman. Sacks was convinced that
ail and that, if something mattered
Goffman's Shetland Islanders (s�e � off. man, 1959): how are they to infer be observable. , it should
�oral character from potentially mlsleadmg appearances?. For instance, in a fascinating passa
sociology of G.H . Me ad 's propo ge, Sacks noted the ba ul influe
sal that we need to study lef nce on
things which are
352

353
DAVID SILVERMAN " POSTSCRIPT
WHO CARES ABO UT 'EXP ERIE
NCE '?

not available to observation, e.g. 'society', 'attitudes' (see Mead, 1934). As Sacks My research is about conversati
on only in thi8 Inc .
' Ide ntal way, that we can get the
comments: actual happenings of on tape and
t ranscn'.be them more or less an
;
.
somet11Ing d therefore have
to begin with If you .
't d a1 Ith the actual det' ail of
But social activities are observable, you can see them all around you, and you can
write them down. The tape recorder is important, but a lot of this can be done
then you can't have a sci�nce of
s :: :
l li e . ( ac1(s, 1 992, VoL 2: 26)
actual events

without a tape recorder. If you think you can see it, that means we can build an Tap es and transcripts als o offer mor "
the first place, they are a ublic r ecor� than .Just something to begin with'. In
observational study. (1992, 1 : 28)
in a way tha �
. t field n 0 tes are not
, avaIlable to the scientific communit
. . . Sec ond y'
they can be rep I aye d and trans-
Despite ethnographers' attention to the logic of writing their field notes (see cnptlOns can be improved and anaI
Emerson et al., 1995), most do not confront fully the problematic character of by the original transcript. As Sack s Yt seS t ke, Off on a different tack unlimited
how we describe our observations. Put at its simplest, this relates to what old h�IS stud ents:
categories we use. As Sacks says: I started to play around with
tape recorded conver
1I cou.ld ty e them out
that I could replay them ' that . satr" ons, for the SIngle virtue
.

Suppose you're an anthropologist or sociologist standing somewhere. You extendedly, who knew 'how � somewhat, and study them
� \ ����
see somebody do some action, and you see it to be some activity. How can you go O ' t Ight take. . . .'It :va sn' t
interest in language, or from . fro m any lar ge
about formulating who is it that did it, for the purposes of your report? Can you so e h trcal formulation of what
� tudi ed, but sim ply by virtue of that; I co . should be
use at least what you might take to be the most conservative formulation -his name? It again and again . And aIso, .
get my hands on It, and I could study
(1: 467) con sequentially others could
studied, and make of it what the ' 10 0k at what I had
y could, if they wa nted to dIsagree with me. (Sacks
1992, VoL 1: 622) ,
As Sacks suggests, this apparently trivial problem is actually not resoluble by
better technique, like detailed note taking. Rather it raises basic analytic issues: A third advantage of detailed transcrip
sequences of utterances withou ts is thatd if you want to, you can ms
. pec
. t
. _

The problem of strategy . . . may not be readily handleable by taking the best notes t b
first researcher. For it is within these '
em g l '
Im lte to the extracts chosen by the
possible at the time and making your decisions afterwards. For one, there is an issue talk, that we make sense of conversa �equences, rathe� than in single turns of
of when it is for the Members that it turns out who did the thing. (1: 468) tIOn. As Sacks pomts ou t:
having available for any given
In fact, many contemporary ethnographers, now aided by advanced software utterance other utterances aro
important for determining wh und it, is extremely
.
at was said. If you have aVaIlab
packages, ignore this problem. In the way Sacks suggests, they simply put in talk that you're now transcribin , . le only the snatch of
g, you re In tough shape for det
some set of categories derived from lay usage 0 : 629). By doing so, of course, (1: 729) ermining what it is.
we are no wiser of how, in situ, categories are actually deployed and enforced,
nor how violations in category-use are actually recognized (1: 635-6). This is why Sacks worked with Iong
As he put it: sequences of tape-recorded interactio
n.
TAP I N G I NTERACTION the kind of phenomena I deal wit
h are always tra nscnph
" ons of actual
in their actual sequence. (Sacks occurrences
, 198 4: 25)
As we have seen, both Goffman and the early Chicago School ethnographers
had generally relied on recording their observations through field notes. Why Sacks liked to work with such material
the machinery through which because h� c�uld use It. to try to access
they try to make sense of each ot:er, P Ie p ut theIr mteractio ns together as
did Sacks prefer to use an audio recorder? eo
Sacks's answer is that we cannot rely on our recollections of conversations. s actIOns. E the activity of 'cod'mg .
Certainly, depending on our memories, we can usually summarize what �ot the preserve of research scientists. All of us ven 'c0d e' what we hear and see
' ls
.
different people said. But it is simply impossible to remember (or even to
note at the time) such matters as pauses, overlaps, inbreaths and the like. Think
m the wo rld around us Thisr:
wh en the y say tha t societal i e:. at � �
G�.r mkel 09 6�) a�d Sacks (1992) me
. SCIe an
observable and reportable. ers , e socral ntIsts, make the world
back to Extracts 1-3 above for a good example of this. 1

Now whether you think these kinds of things are important will depend
upon what you can show with or without them. Indeed, you may not even
Put at its simplest, this means that re
they use categories. For instan �earc1lers must be very careful how
ce S a
appear to have no problem in char�ct .s quote.s from. two linguists wh o
�
be convinced that conversation itself is a particularly interesting topic. But, at enzl�g partIcular (mvented) utterance
as
.Iza'sim
least by studying tapes of conversations, you are able to focus on the 'actual .tlOnspleof'' 'complex', 'casual' or ,ceremomal' . For Sacks, such rapId . charac s
ter-
details' of one aspect of social life. As Sacks put it: data assume
354
355
 WHO CARES ABOUT 'EXPERIENCE'?
DAViD SILVERMAN (f POSTSCRIPT

accurate] without an analysis of what it categorize (1: 803). Only in this way can sociology be 'a natural observational
that we can know that [such categories are science' (1 : 802).
1: 429)
is [members] are doing. (Sacks, 1992, Vol. Let us take a concrete example. In two of Sacks's lectures, he refers to a
might . respond th�t Sacks has
At this point, the experienced researcherover
New York Times story about an interview with a navy pilot about his missions
as
characterized conventional research assu -naIve. In partlcular, most in the Vietnam War (1: 205-22, 306-11). Sacks is specially interested in the
ming any one-to-one correspon­
researchers are aware of the danger ofaspects
story's report of the navy pilot's reported answer to a question in the extract
'reality' which they �urport
derKe between their categories and the (1949),ofman
below:
y researchers claIm that
to describe. Instead, following W eber tructs (or 'ideal types') whI. ch are only
they are simply using hypothetical cons are useful, not whether they are THE NAVY PILOT STORY
to be judg ed in relation to whether they
'accurate' or 'true '. notes: Ho� �id he feel about knowing that even with all the care he took in aiming only
However, Sacks was aware of this argument. As he at mlhtary targets someone was probably being killed by�his bombs?
in the s�cial sciences. t� �ropo.se that the
It is a very conventional way to proceed
machinery you use to analyze som e da.ta you
� � � �
have IS acceptab e If It IS n?t l ten �d y 'I certainly don't like the idea that I might be killing anybody,' he replied.

the analysis of real phenomena . That is,

�
you can h ve machmery whIch IS a vdhd 'But I don't lose any sleep over it. You have to be impersonal in this
business. Over North Vietnam I condition myself to think that I'm a
somethmg for you. (1: 315)
hypothetical construct', and it can analyze military man being shot at by another military man like myself.'
Source: Sacks, 1992, Vol. ] : 205
By contrast, the 'machinery' in wh.ichanSa ��s is in�ere�ted is notouta �t�et be of
'hypothetical constructs'. Instead, hIS �.bltl O � s cla;m IS thro ugh
ld' (1: 316) . The mac hme ry he sets out, then , IS. �ot Sacks invites us to see how the pilot's immediate reply ('I certainly don't
dealing with the real wor but the actual categones
to be seen as a set of more or less useful categories like the idea . . . ') shows his commitment to the evaluational scheme offered
and mechanisms that members use.4 by the journalist's question. For instance, if the pilot had instead said 'Why
do you ask?', he would have shown that he did not necessarily subscribe to
the same moral universe as the reporter (and, by implication, the readers of
A TEXTUAL EXAMPLE the article) (1 : 211).
Having accepted this moral schema, Sacks shows how the pilot now
We saw earlier an example of the kindeme of work on conversation that Sacks �u�l�s an ans",;,er which helps us to �ee him �n a favourable light. The category
inspired in Don Zimmerman's study of Sack rgency telephone calls. .However, mIlItary man works to defend hIS bombmg as a category-bound activity
it would be misleading to assume that s's work is only of mterest to which reminds us that this is, after all, what military pilots do. The effect
of this is magnified by the pilot's identification of his co-participant as 'another
researchers who want to study talk. g data of all kinds. . military man like myself. In this way, the pilot creates a pair (military man/
Sacks was interested in naturally occurrintape or textual mat enal
.
, ac��rdmg military man) with recognizable mutual obligations (bombing/ shooting at
Whether we are look ing at aud io or vide o
to Sacks we can still do good analyses prov iding we study how partlCIpants the other). In terms of this pair, the other party cannot properly complain or,
(membe�s) 'put some category in' (in diffe rent ways, this is the c?ncern shared as Sacks puts it:
by all the authors whose chapters app.eard in this boo k). In tl:lS respect, for
Sacks, the discipline of sociology has faIle to grasp
the analytIc nettle. As he �
there are no compla nts to be offered on their part about the error of his ways, except
If he happens to VIOlate the norms that, given the device used, are operative.
puts it: (1: 206)
� �
the sense that the simp.ly ut .some
All the sociology we read is unanalytic, in Notice also that the pilot suggests 'you have to be impersonal in this business'.
in doin g that, but they re domg It slmply
category in. They may make sense to us Note how the category 'this business' sets up the terrain on which the specific
, my emphasls)
as another Member. (Sacks, 1992, Vol. 1 : 41-2 pair of military men will shortly be used. So this account could be offered by
The availability of alternative category collections (e.g. first na�es, surnames, either pair-part.
occupational or family titles) means that any sociology that .asplfes to do more However, as Sacks argues, the implication is that 'this business' is one of
than reiterate commonsense understandings must speCIfy how members many where impersonality is required. For
357
356
DAVID SILVERMAN POSTSCRIPT
WHO CAFI ES ABOU T 'EXP ERIE NCE'
?
if it were the case that, that you had to be impersonal in this business held only for SO WHAT?
this business, then it might be that doing this business would be wrong in the first
instance. (1: 206) On the face of it, teachers of soc p rob
professionals would seem to b verialy unl . leymsauda �d, even more so, caring
Ikel
Moreover, the impersonality involved is of a special sort. Sacks points out that work. Indeed Sacks refused to::�po t� sug Ience to appreciate Sacks's
we hear the pilot as saying not that it is unfortunate that he cannot kill justify itself b� its contribution to sone�y , �� ns �hat social science must
� stio
'personally', but rather that being involved in this 'business' means that one with a student shows: as t IS questIOn and answer session
must not consider that one is killing persons (1: 209).
However, the pilot is only proposing a pair of military man/military man. Q: Will it e�er have any pos
sibl e rele van ce to the peo ple
In that sense, he is inviting the North Vietnamese to 'play the game' in the Invol ved In prod ucin g it? Who wer e
same way as a child might say to another 'I'll be third base'. However, as Sacks A: It nee dn 't have any rele van
ce . . . I take it that ther e's an
notes, in children's baseball, such proposals can be rejected: amo � nt of stud ies
.
that are not inte nde d to be rele
eno rmo us
stud ies of how can ce r d oes van t . For exam p I e,
can cer IS .
' not Inte nde d to bUil d bett er
if you say 'I'll be third base', unless someone else says 'and I'll be . . . ' another canc er.
position, and the others say they'll be the other positions, then you're not that thing. (Sac ks, 1 992 , Vol. 1 : 470 )
You can't play. (1: 307)
In this answer Sacks 1:seIy f 11 s what
Of course, the North Vietnamese indeed did reject the pilot's proposal. Instead, ological indiff�rence' � ques t�IOnow Ga�finkel terms 'ethnom od-
s 0f h ow thmgs (whether cancer eth . 1
they proposed the identification of the pilot as a 'criminal' and defined .
P:oblems or, mdeed, social science itself) should be SOC Ia
As I have alrea' dy noted, '
themselves as 'doing police action'. hIS answer depends 0 de S . ' sum
- pre
As Sacks notes, these competing definitions had implications which
went beyond mere propaganda. For instance, if the navy pilot were shot
not intend that researc� �ho���l�l �!n �
However, we do not hav· e to J'udge Sacks ' s
� :: ?
ably the stud ent did
. contrib
'
P roblems by thOIS answer. In partIcu potentIal ution to social
down then the Geneva Conventions about his subsequent treatment would lar, r � cen t ap l'Ie d . wo rk usin
only properly be applied if he indeed were a 'military man' rather than a that ultimately derive from Sacks 1 as, m r: g methods
buted to the solution f �
'CtI . my VIew , und oub tedl y contri­
institutional interactio� (:::;�:;if:ge, thIS volum o
'criminal' (1: 307). a C � l lss ues, m ost nota bly in
Unlike more formalistic accounts of action (Parsons, 1937; Mead, 1934), the field of
Pera"kyIa," 1995; Silverman, 1997; Hea e; rew and Heritage, 1992b;
Sacks's analysis shows us the nitty-gritty mechanisms through which we
construct moral universes 'involving appropriate kinds of action and par­ The practical relevance of such 'appthlied and Luff 2000)
' conv�rsatio� ana l sis (CA sh
, : u a ';
:
(��� � �J l��� . �l� ;� t a� : S the SOCiology of s�ci�l pr�ble=�
ticular actors with motives, desires, feelings, aspirations and sense of justice' er on atte to
(J. Gubrium, personal communication). Like Garfinkel (1967), Sacks wanted
to avoid treating people as 'cultural dopes', representing the world in ways
terms of social science concepts like 'role' 'lab l i o
!
precisely becau�e it �vo�ds :�; ':bft��� f�� �� �;d ;:��; ca� be usef�l
a ctl�n (e. � . m
that some culture demanded. Instead, Sacks approached 'culture' as an
'inference-making machine': a descriptive apparatus, administered and used and 'Societal reactions' set careers of devia�ce � � el' 'd '
vlance ') ' S o, lf. labellmg'
.
in specific contexts. Although their language is different, the chapters by . g trou in otlOn, such mobon depends
upon 'the ordered acti " tIes 0f tellm
. :�
Such ordered actIVItIes can be made VIS bles and proposing problems' (325) .
Atkinson and Coffey and by Prior in this volume imply much the same . 'bl e beca
CA demands 1 use, as Maynard puts it,
approach to description.
Sacks's take on 'culture' underlines a radical position that may be unsettling [the] observation and scrutiny of the
to some readers. After all, many of us are wedded to particular methods and details of actual talk and interaction.
(320)
approaches and are unlikely to be prepared to abandon them overnight. So
lu
why take account of what Sacks had to say?
I will try to answer such an understandable response in two ways. Shortly, �����:�:��I S���:�;���:::�;���f:;s���C�::�i��st:ion to the study
I will summarize what I take to be Sacks's contribution to 'mainstream' quali­
' er� CtI. ?na� phenomena as dive
To the extent that we have learned abo
tative research. First, however, I will address the repeated criticism that ut such lI
Sacks's position (and, more generally, ethnomethodology) leads to an arid,
conflict, domination, troubles and p
ance, and mediated verSI. Ons 0f S .
robl ems ' � rsity,
, t e mstrtuhonal pro cessing of dev
i-
OCIal problems it is because we have prop
purely theoretical take on institutions which glories in its lack of substantive about systems of vernacular talk ' ositions
input into social problems. I will call this the 'so what?' line of criticism. knowledge, conversational sequen�in
t
��%: ��� �;�:�� ����;
t k'
to
� -
.
g, commonsense

358
359
 a

WHO CARES ABOUT 'EXPERIENCE'?

DAV ID SILV ERM AN POSTSCRIPT

.. the gap between beliefs and action and between what people say and what
J

. . at Sac k s sal'd ab ou · t tur n-t aki ng that is rel. peva nt they do (Webb and Stimson, 1976; Gilbert and Mulkay, 1984).
It '
IS no t Jus t wh
Moreover, rary studIe. s of SOCIa. I pr blems Sacks' s work on 'descn erse' t'
IOn
to contempoa generati n of ethn�g�ap� who have studied topics as div yed Qualitative researchers' preference for interview studies ironically respects
has inspired ess hean?ngs (�Iolste1l1, 1 ;��) schemes to get the unemplocare a division of labour preferred by quantitative researchers. According to this
as mental illni11er, 1993), fanulythe a y (Miller 1987; Gubrium, 1992), the division, while quantitative research focuses on objective structures, it falls
into work (My (Gubrium, 1980) an� :us'ness �eetings (Boden, 1994). gh to qualitative researchers to give 'insight' into people's subjective states.
of the elderl int, the reader may s11' 11 as�.' �here does this get me? Althou I am The unfortunate consequence of this division of labour is that both approaches
At this po y tra dition of ethnograp h'lC work that follows Sacks, . neglect a great deal about how people interact. Put more strongly, both kinds
a 1 Ive
· I
there is many readers are unlI'k eIy t nt to abandon more ma tion
0
1l1s tre am of research are fundamentally concerned with the environment around the
that
aware ic work in order to Iearn or membership cate ;; goriza phenomenon rather than the phenomenon itself.
ethnograph Moreover, we need to question the argument that observational or other
analysis.
me con clu de , the ref ore , bY und
. erl
. ini ng the implications that Sacks's naturally occurring data are 'unavailable' in the supposedly 'private' sphere
Let earch.
work contains for 'mainstream' quahtatJve res of human interaction (e.g. in domestic life). As Gubrium and Holstein have
noted:
i M PL ICATIO NS The formulations of domestic order that we hear outside households are treated

tat ive res ear che rs are ? eco m in inc rea sin gly sensitive to assedessiningsomthee as authentic as are those heard within them. . . . As a practical matter, this means
Quali I py work is still accept
work. Even. lf2�����b1) �qs:It
that the analyst would treat private and public as experiential categories -- constructed
quality of t�eir rm an , 200 0. ity' issues are central t� many and oriented to by interacting persons not actual geographic or social locations.
journa ls (SI lve y tex tbo o k s ' ( se e P e rak yla , this volum e; MIle s and . . . [This implies that) methodologically, we should not take for granted that privacy
quali tat ive me tho do log le et al ., 20O4)' s. debate can be summanz. ed
implies privileged access - that those occupying the private sphere are taken to be
Huberman, 198st4h,' um SeaIe, 1999"' Sea 0 t thl
experts on its description, the final arbiters of its meaning. (1995: 205)
Sacks's (po ISs . ou. s) contn. butlf' On
0 ou r me tho ds , the quality of aoubelr dat a and
in ter ms of thr ee ue s. the qu alIt y
. . I WI'n briefly deal with each are ow . Such a situation suggests that we need to look twice at the unthinking
the qu ali ty of ou r da ta an al Y SIS identification of the open-ended interview as the 'gold standard' of qualitative
research. Note that this is not to reject each and every interview study. I merely
suggest that the choice of any research instrument needs to be defended and
Qu ality of methods
els ew he re �he wh ich int erv iew s are by far the most. that the pursuit of people's 'experience' by no means constitutes an adequate
I have shownthod used 11 q::��i��t��e research (Silverman, 2000: 290-2)er defence for the use of the open-ended interview.
common mecertainly w uld 1 n t Wl. s11 t ar ue that interviews are nev
Althoug.h the romantic auspI?ces f so;e i�erview studies should give us
I ? 0
appropnate, ught. Quality of data

pause for thoiews, as Hent. age puts 1. t, the mistake is to treat the verbal Whatever material Sacks analysed, he was scrupulous about working with
In interv long sequences of text or transcript and providing those sequences intact to
formulations of subjects the reader. By this means, he provided a public forum in which to assess the
. for the observation of actual
ge,
behaviour. (Herita reliability of his data.
as an a ppropnate substitute Of course, data quality is also something that concerns 'mainstream'
1984: 236)
ect impact on thedatkin d qualitative researchers. Yet how often are textbook discussions translated into
d He rit ag e (19 92a ) s 11�w ho w thi s ha s a dir h a as practice? How often are we given details about the auspices of field note
Drew an think are relevant. ost qualitative researchers use suc the extracts? How frequently do interview studies provide sequences of talk
of data. we s groups and dla. ne. s. They thus attempt 'to get inside and (including that of the interviewer)?5
interVIeWS, fofocuSOCial
. .
111s . tlO' ns t� gal'n access to their interior proces: ses
tItu Of course, some of this negligence can be ascribed to the ever-increasing
x"
"black bo, However, such studIes may suffer from two problems demands of journal editors for shorter research articles. How can we satisfy
· (5) ·
practIces whI. ch these pragmatic constraints and yet provide credible data extracts? Using
con tex t ( e.g . th e ' or ga nization') to
.. the assumppot·Ionndof; a stable reality or Sacks's work as our example, we might use fewer but more detailed extracts.
people res 361

360
 &
DAVID SILVERMAN POSTSCRIPT
WHO CAR ES ABO UT 'EXPE
RIEN CE'?
As I tell my research students, saying 'a lot about a little' is not a bad maxim an aspiration of quite con
for qualitative research. provIde a shining example.ventional 'mainstream' qualIta . re
' tIve search and to
Quality of data analysis
C O NCL U DI NG R E
MARKS
Analysing long, well-transcribed data extracts, as Sacks did, helps both the
reliability and likely validity of our data analysis. Sacks was no revolutionary Sacks's endeavour ultimately
in this respect. He merely followed the path prescribed since the 1950s at least he himself put it: reSI.' sts any formulation which I
might offer. As
through the method of analytic induction (AI) (see Denzin, 1989: ch. 7).
J can tell you som
AI involves the use of the constant comparative method allied to deviant
case analysis. Yet, only rarely is it explicitly applied in mainstream ethnog­
ething bu t
upshot of what I've said is h tlJ�� �: �
� t be caref�l what you make
of it. . . . The
raphy (see Bloor (1978) for a rare example). More commonly, there is an anyone makes of wha t it is e 0 comrmtment to what
that I d o nor tO whatever kind of placing
anecdotal quality to many published qualitative research studies, first noted might provide me, which
turn on suc'h a treatment. (19 recom mendations anyone
by Mehan (1979: 15). This problem is succinctly expressed by Bryman: 92, 1: 621 )
Like the work of Wittg
largely available to us inennosteteb
in and Saussure before lU. rn,
There is a tendency towards an anecdotal approach to the use of data in relation to
ook d 1eCtures p ublished Sack s's work is
conclusions or explanations in qualitative research. Brief conversations, snippets Like these other two great thm O kers, sSan p osth
ack s wI. ll cont'mu e to ex ert an umo
' '

usly.
from unstructured interviews . . . are used to provide evidence of a particular on many subsequent generation infl ue nce
contention. There are grounds for disquiet in that the representativeness or Howev:r, although Sacks cans.ins
generality of these fragments is rarely addressed. (1988: 77) do . Followmg Wittgenstein's (1971 pire, he cannot tell us all that we have to
work to climb up higher but w ) ana1ogy of the 'ladder', we can use Sacks's
Bryman's complaint is mirrored ten years later by Paul ten Have. The latter To repeat his words: ' e have to proceed carefully and
thoughtfully.
notes the complaint that in CA, like other kinds of qualitative research:
I can tel l yo u something, bu
findings . . . are based on a subjectively selected, and probably biased, 'sample' of .t you have to be careful wh
at you make of it.
cases that happen to fit the analytic argument. (ten Have, 1 998: ch. 7, p. 8)

This complaint, which amounts to a charge of anecdotalism, can be addressed

by what ten Have, following Mehan (1979), calls 'comprehensive data treat­ * I am most grateful for the com
ment'. This comprehensiveness arises because, in qualitative research, 'all Leydon on an earlier draft.
ments of JudI'th G reen, Jay
Gubrium and Geraldine
cases of data . . . [arel incorporated in the analysis' (Mehan, 1979: 21).
Such comprehensiveness goes beyond what is normally demanded in many 1 . Aldrich also makes the Com
mon assum tion that open
quantitative methods, For instance, in survey research one is usually satisfied .
to all paradigms which re 'ect
) func tiona I:'Ism:
l
-ended ll1t erviews are central
by achieving significant, non-spurious, correlations. So, if nearly all your data . . . understand Its· organization
s as members do.
'to study organ'Izat'IOns an
analyst mu st
.
support your hypothesis, your job is largely done. unstructured ll1t erviews, doing fieldwork (an In pra . .
cfIce, thOIS means conductin
g
d) pa rtICIpant observation'
By contrast, in qualitative research, working with smaller data-sets open Gergen reveals very clearly
"SId es' : 'th th e rom . .
antICIst aus' pices 0f AldTIc (19 92: 25) .
to repeated inspection, you should not be satisfied until your generalization e chief contribution of the ' h , s language of
rom antrclsts to the preval'1'111g con
, .
creation of the deep mterz.or
person was their rhetorical
is able to apply to every single gobbet of relevant data you have collected. of capaci. tIes
'
. cept of the
lying deePIY wlth . , . . the eXIStence 0f a repository
' or characteristics
The outcome is a generalization which can be every bit as valid as a
'

208-9). ll1 human consciousness' (199

statistical correlation. As Mehan puts it: Romantic accounts can be
2:
. seen, as Gergen suggests,
111 themselves derived from as
The result is an integrated, precise model that comprehensively describes a specific . wha t IS' the case (and) ach . 'forms of language' not
rh etoTIc al artifice' (1992: 210) Al ' ll1g th elr
IeV " Impact through
phenomena [sic], instead of a simple correlational statement about antecedent and modernist direction, we cou ' though Gergen takes this argument in a post-
' 0d01og y's focus on the
consequent conditions. (1979: 21) ld use I' t to support ethnom
?roperties of actual language use (e.g . In . eth
Implied in Gergen's discus members' descriptio ns) . Ind
eed, this is'
To anyone who cares to read them, Sacks's own data analyses (Sacks, 1992) 2 . S'll1ce the authors of both
sion of 'representation as
a communa l art
c
' ct' (21 4).
Ifa
are a model of comprehensive data treatment. His work serves to underline these stUdIes
' refer favourably
to be very ungracious in usi to my own work, I may seem
ng their work as examples
of a consensus which I reject
.
362

363
DAVID SILVERMAN POSTSCRIPT

WHO CARE S ABO UT

. . 'EXP ERIE NCE'?
Let me emph aSlze that I believe that theIr worl( IS ' good of its kind, unlike some Atkinson, P. and Silverm
. an, D. ( 997) Kundera's
eaSIer targets. and the Invention of Im
.
3. A hidden aspect of thIS consensus is that data gathered are under-analysed. A focus Self. Qualitative Inqu iry, 3 mortality: The Inter view Society
. Bloor, M. (1 978) On the (3) : 304-25.
on 'themes' can concea1 rese an
. �rchers' preference f0r 'telling' examples, l.e. examples and uses of inductive alysis of observational da ta : a discus sion of
. � � ;
.
that support their case (se� S:lvern:an, 20O ' es eciall 283-96). Moreover, very l'ttle
attention 1S usuaIIy pcaid 111 111tervlew stu ies o ho interviewer and intervIewee
1
545-57.
techniq ues and respon
dent validation . Sociolo
the w orth

��
�o-construct agreed meanings (see {a 1ey� 2004) '
Boden, D. (1 994) The B
usiness of Talk: Organ
gy, 1 2 (3) :

� � ;� ;� �� ;�
Press. izations in Action. C
4. Go odwin and Heritage ( 990 3� 'the great debate between cognitive ambridge: Polity
anthropology and cu1tu�a n:a er li
. �
e status of emic analysis'. The d bate
Bryman, A (1 988) Qua
Clavarino, A, Najman, J.
ntity and Quality in Soci
al Res
and Silverman, D. (1995) earch. London: Unwin Hyman.
turns on the relative rehablhty 0 nat've
I and analysts' understandings of cu tur� .
] (2) : 223-42.
data. Qualitative inqu iry, AsseSSing the quality of
qualitative
��:��� � :
However, as both Garfinkel and S ack how commonsense and analytic
understandings are in no waY com t , e' I s ead ;he point is to see how mem?ers
. Denzin, N.K. (1989) T
he Research Act (Third
t confusi�g commonsense interpretatlons
Hall. E dition). Englewood
produce the sense that the� d 0 w Cliffs, NJ: Pren tice
D enzin, N.K. and Lin
with social science explanations. coln, Y. (2000) Handbook
. by Thousand Oaks, CA: of Qualitative Research
These imperatives have been taken �n board in two major ethnographIes Sage. , (Secon d Edition) .
Drew, P. and Heritag
anthropologists (Moerman, 1988; �oodw111 1990) As Goodwin and Heritage argue,
�nts 'but instead rellyI upon the actions
e, J. (199221) Analyzing
��;� �� t���
Heritage (editors), Talk Talk at Work. In Paul
these writers do not appea: t ,
� hey interactively and sequentially make
at Work. Cambridge: C Drew a
S Drew, P. and Heritage,
J. 0992b) (eds) Talk at ambridge University Pres nd John
of participants themselves ( . l Work. Cambridge: C ambr s, 3-65 .
Press. idge University
sense. Eberle, T. and Maeder, C. (
In terms of my argument above, we might call such stud'les examples of non- 2002) 'The goals of the con
romantic ethnograp hY·
Conference on Ethnog feren
raphic Organization Stu ce', Conference Proceedings,
5. Neither of the organizatl. 0nal th g hies discussed earher
: . SWitzerland. dies, University of St
Em erson, R, Fretz, R
�� ::� �
(Hassard , 1 993,' Gallen,
Linstead and Thomas, 2oo2) gi e thing more than brief extracts of wha and Shaw, L. (1995) Writi
. . Chicago University Pre ng Ethnographic Fieldwork.
Flick, U . (1998) An Introdu
intervIewees sal' d . We therefore get no 111 formation c on how these responses ar ss. Chicago:
.
Garfinkel, H (1967) Studies
contextualized within an interaction. ction to Qualitative Resea
rch. London: Sage.
in
Gergen, K. (1992) Organiz E thnom ethodology. Englewood Cliffs, NJ: Pre
ation ntice Hall.
Hughes (eds), Rethinking O Theory in the Postmodern Era. In M. Reed
London: Sage, 207-26. rganization: Directions in and M .
Organization Theory and
Recommended Gilbert, G.N . and M ulka Anal ysis.
y, M. (1984) Open ing P
Scientists' Discourse. C andora's box: A SOCiologi
ambridge: Cambridge cal Analysis of
C offman, E. (1959) The University Press.
Presentation of Self in E
Anchor. veryday Life. New Y
ork: Doubleday
Goffman, E. (1 981) Forms
of Talk. Oxford: Blackw
GoodWin, M .H (1990) ell.
He-Sa id-She-Said: Talk as
Bloomington: Indiana Un Social O rganization am
iversity Press. ong Black Children.
GoodWin, C. and Her
itage, J. (1990) Convers
Anth ropology, 1 9: 283-3 ation AnalYSis. Ann ual
07. Review of
Gubrium, J. (1980) Pati
ent Exclusion in Geriatri
21: 335-48. c Settings. The Sociolo
gical Quarterly,
Gubrium, J. (1992) Out of
Control. Newbury Park
Gubrium, J. and H olste CA: Sage.
in, ]. (1995) Qualitative
Experience. Qualitative Inq Inquiry and the D epriv
R EFEREN CES Gubrium, J, and Hol uiry, 1 (2) : 204-22. atization o f
stein, J. (2002) Handbook
Aldrich, HE. (1992) Incommensur b
of Interview Resea rch.
� � r; 7
CA: Sage. Thousand Oaks,
eth nki�g Organization: New Directions In
di ms? Vital signs from three perspectives. Hammersley, M. (1989)
In M. Reed and M. Hughes e s The Dilemma of Qualitat
;� �
' Tradition . London: ive Method: Herbert Blu
�
m er a nd the Chicago
Has sard, J. (I 993) Soci
e . ti�g t e Relationship between partlClp
Organization Theory and Resea . ond on' Sa e, 1 7-45. Rou tledge.
Atkinson, P. and Coffey, A (2? ;;
. , , n ology and O rganization
t�
Observation and IntervleW1l1g. In ' ubrium and J , Holstein (eds), Handboo c 0
Postmodernity. Cambr
ten Have, P. (1998)
idge: Cambridge Unive
Theory: Positivism, P
rsity Press.
aradigms and
Interview Research. Thousand Oaks, CA.. S age, 801-14. Doing Conversa tion A
Qualitative M ethods S nalys is: A Practical
eries) , London : Sage. Guide. (IntrodUcing
364

365
;:1 . ..

DAVID SILVERMAN POSTSCRIPT

WHO CAR ES ABO UT 'EXP
E R IENC E'?
Heath, C. and Luff, P. (2000) Technology and Action. Cambridge: Cambridge University "':ittgenstein, L. (971) Tra
Press. ctatus Lo?ico-Phz'losophz.
ZImmerman, D. (1992) . cus. London: Routledg
Heritage, J. (1984) Garfinkel and Ethnomethodology. Cambridge: Polity Press. The Interachonal organ ' . e.
Holstein, J. (1988) Court ordered incompetence: conversational organization in �� � � �
I re nd J. Heritage (
I �, 1 6 . eds), Talk �����:� �� �
at Wo
lls f r eme gency s rvi
r ge.. ambndge U : ces.
�
involuntary commitment hearings. Social Problems, 35 (4): 458-73. mversity
Hume, M. (2002) These mawkish tears are an insult. The Times, 19 August.
James, H. (n.d.) The Portrait of a Lady. London: Heron Books.
Linstead, A. and Thomas, R. (2002) 'What do you want from me'?' A Poststructuralist
Feminist Reading of Middle Managers' Identities. Culture and Organization, 8 (1):
1-20.
Maynard, D.W. (1988) Language, Interaction and Social Problems. Social Problems,
35 (4): 311-34.
Mead, G.H. (1934) Mind. Self and Society. Chicago: Chicago University Press.
Mehan, H. (1979) Learning Lessons: Social Organization in the Classroom. Cambridge, MA:
Harvard University Press.
Miles, M. and Huberman, A. (1984) Qualitati'ae Data Analysis. London: Sage.
Miller, G. (1987) Producing Family Problems: Organization and Uses of the Family
Perspective and Rhetoric in Family Therapy. Symbolic Interaction, 10: 245-65.
Miller, G. (1993) Enforcing the Work Ethic. Albany, NY: SUNY Press.
Moerman, M. (1988) Tallcing Culture: Ethnography and Conversation Analysis.
Philadelphia: University of Pennsylvania Press.
Parsons, T. (1937) The Structure of Social Action. New York: McGraw-Hill.
Perakyla, A. (1995) AIDS Counselling. Cambridge: Cambridge University Press.
Rapley, T. (2004) The Qualitative Interview. In C. Seale, G. Gobo, J. Gubrium and
D. Silverman (eds), Qualitative Research Practice. London: Sage.
Sacks, H. (1963) Sociological Description. Berkeley Journal of Sociology, 8: 1--16.
Sacks, H. (1972) Notes on Police Assessment of Moral Character. In D. Sudnow (ed.),
Studies in Social Interaction. New York: Free Press, 280-93.
Sacks, H. (1984) Notes on methodology. In J.M. Atkinson and J. Heritage (eds),
Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge
University Press, 21-7.
Sacks, H. (1992) Lectures on Conversation, edited by Gail Jefferson with an Introduction
by Emmanuel Schegloff. Oxford: Blackwell, 2 vols.
Seale, C. (1999) The Quality of Qualitative Research (Introducing Qualitative Methods
Series). London: Sage.
Seale, c., Gobo, G., Gubrium, J. and Silverman, D. (eds) (2004) Qualitative Research
Practice. London: Sage.
Silverman, D. (1989) The Impossible Dreams of Reformism and Romanticism. In
J. Gubrium and D. Silverman (eds), The Politics of Field Research: Sociology Beyond
Enlightenment. London: Sage, 30-48.
Silverman, D. (1997) Discourses of Counselling: HIV counselling as social interaction.
London: Sage.
Silverman, D. (1998) Harvey Sacks and Conversation Analysis (Polity Key Contemporary
Thinkers Series). Cambridge: Polity Press; New York: Oxford University Press.
Silverman, D. (2000) Doing Qualitative Research. London: Sage.
Sudnow, D. (1965) Normal crimes: sociological features of the penal code in a public
defender's office. Social Problems, 12: 255-76.
Webb, B. and Stimson, G. (1976) People's Accounts of Medical Encounters. In
M. Wadsworth (editor), Everyday Medical Life. London: Martin Robertson.
Weber, M. (1949) Methodology of the Social Sciences. New York: Free Press.

366

367
 APPEND IX
Underlining indicates stre
ss or emphasis.
A 'hat' �r CIrcumflex acc
ent symbol indicates a ma
rked p itch ns. e.
.

Appendix Equal sIgns (ordinarily at ,

an ensuing one) indicate a the end �ne h �e and the start of
Transcription conventions between them. t
'latched' r� atIonship no
silence at all
Fuller glossaries m ay be' f . Sa
( 9 74) 'A slmp
' lest systemati
ound In cks, Fl., Scheg! off. , E.. A. and
for th e organ'c Jefferson G .
aking 'for
Conversa tlO '
' n', Language 50: 696-735cs Iza f JOn of tu
1
. rn-t
The examples printed embody an effort to have the spelling of the words
( � 984) �tructu res of Soci;Z A
Cambndge U niversity Pre c tion StU
:
' �
d��\ �ok Ins�
n, J:M and Heritage, ]. (eds)
nversatlOn A nalys is. C
roughly indicate how the words were produced. Often this involves a ss. ambridge:
departure from standard orthography. Otherwise:
--> Arrows in the margin point to the Jines of transcript relevant to
the point being made in the text.
() Empty parentheses indicate talk too obscure to transcribe. Words
or letters inside such parentheses indicate the transcriber's best
estimate of what is being said.
hhh The letter 'h' is used to indicate hearable aspiration, its length
roughly proportional to the number of 'h's. If preceded by a
dot, the aspiration is an in-breath. Aspiration internal to a word is
enclosed in parentheses. Otherwise 'h's may indicate anything
from ordinary breathing to sighing to laughing, etc.
Left-side brackets indicate where overlapping talk begins.
Right-side brackets indicate where overlapping talk ends, or marks
alignments within a continuing stream of overlapping talk.
o
Talk appearing within degree signs is lower in volume relative to
surrounding talk.
> < 'Greater than' and 'less than' symbols enclose talk that is noticeably
faster than the surrounding talk.
((looks» Words in double parentheses indicate transcriber's comments, not
transcriptions.
(0.8) Numbers in parentheses indicate periods of silence, in tenths of
a second - a dot inside parentheses indicates a pause of less than
0.2 seconds.
Colons indicate a lengthening of the sound just preceding them,
proportional to the number of colons.
becau- A hyphen indicates an abrupt cut-off or self-interruption of the
sound in progress indicated by the preceding letter(s) (the example
here represents a self-interrupted 'because').

369
 £

NAM E IND EX
Danet, B. 103
Fehr, B. 300n
Darb o-Peschanski, C.
Darnton, R 78, 92.
12. Feldman, M. 58 272--3, 275, 278 _80,
Feldman, R 311 287-8, 364n
Name Darwin, C. 78
Das, R. 180
Goo dwin, M. 223, 267,
Finch, J. 181 272-3,
279-80, 28 7, 364n
Gorden, R 141
Fine, G. 128
Davies, P. 311
Dawson, L. 12.6, 12.9 Fine, M. 181
Finlay, A 322
Gou ld, S. 78
Firth, A 287
Dean, D. 260 Grace, A. 202
Fisch, R 335
Delamont, S. 31
Denzin, N. 35-6, 127,
Greatbatch, D. 238, 278
12.9, Green, Jf 7, 363
Greenba um, T. 1 77
Bers, T. 181
Fisher, S. 239
Flick, U. 343
343, 362.
Ainsworth-Vaughn, N. 43 Carey, M. 177, 1 83·-4
Desrosieres, A. 28n Griaule, M. 29n
Agar, M. 182, 187 Bessy, C. 1 1 , 23, 31n Carmines, E. 294 Fonkert, D. 313 , 322
Fontana, A. 144
DeVault, M. 143 , 152., Griesemer, ]. 79
156
Bijker, W. 78
Alasuutari, P. 1 42, 295 Best, S. 36 Cartier-Bresson, H. 253
Deverell, K. 311 Gubrium, J. 4-·5, 7, 40,
Albert, E. 226 Cartwright, 1. 262n Foucault, M. 3, 35- 7, 46, 51,
Dickens, C. 89 39-40, 140--4, 1 49-152, 155·_8,
Birdwhistell, R. 267
Aldrich, H. 343-5, 347, 363n Billig, M. 202-4, 216 Carvajal, D. 1 1 2 44-5 1, 77, 83, 1 13, 149,
Dibbell, ] . 95, 99 162, 1 75, 1 77, 182, 188,
Alkula, T. 289
Alexander, V. 262n Casey, M. 179, 196 201·-2, 262, 31 8
Diderot, D. 78
Dingwall, R 6-7, 39,
Blaxter, M. 184 Certeau, M. de 29n Foweler, F. 143 343, 360-61, 363

Hak, T. 84, 86, 295

Altheide, D. 145, 289 Bloomfield, B. 58
Anderson, N. 30n
Anderson, D. 257 Bloor, M. 6-7, 58, 179, 309,
3 1 3-14, 317, 362
Anspach, R. 339
Blumer, H. 134, 142
Antaki, C. 206 Boden, D. 35, 42, 203, 269,
Arendt, H. 9 300n, 339, 360
Aristotle 79 Bogardus, E. 195n
Boltanski, 1. 16, 19, 22, 26,
Arney, W. 43
Armstrong, D. 318
30n
Ashmore, M . 202 Born, G. 252-4
Atkinson, J.M. 35, 42, 225-6, Bosk, C. 28n, 332-3
238, 268, 272, 289, 291 Bourdieu, P. 26-7
Atkinson, P. 3-4, 31, 58, 202, Bowker, G. 61, 78, 87
284, 292, 309, 344, 358, Bowling, A 82
Atkinson, R. 144
364 Boyd, E. 238, 240, 298-9
Brannen, J. 48
Aubry, J. 177 Braun, V. 187
Austin, J.1. 1 2 Briggs, C. 140-1
Bromseth, J. 102-3, 1 1 8
Champy, J. 325 58, 2.86, Fox, K. 3, 51
Fox, R 15, 20, 28-9n
326, 330, 332, 338
Dodi er, N. 3, 16, 19, 2.6, Halkowski, T. 239
Chaney, D. 61
Chaplin, S. 262 Frankel, M. 1 1 8
Frankel, R 236, 239
30-1 n
Doisneau, R 2.53
Hall, S. 253-4, 262
Charmaz, K. 1 27, 1 29-30
Hamilton, P. 253-4
Chase, S. 49 Frankenstein 77
Dollinger, S. 2.48 Hammer, M. 325
Chateauraynaud, F. 1 1 , 23, Frei dson, E. 314
Frey, f. 144
Douglas, J. 81, 146-8 Hammersley, M. 284-5
30-1n , 292,
Douglas, M. 78, 138 306, 309, 323, 343
Frith, H. 180
Chawla, S. 325
Douglas, S. 1 1 9 Hammett, D. 89
Chen, S. 1 20n, 12.1 Fros t, D. 211
Drew, P . 42., 188, 2.03, Harding, S. 126, 129,
Cherny, L. 100 2.24-6, 142
231, 2.33, 2.35, 237-9 Hassard, ]. 344
Chiapello, E. 30n , 2.69, Gad amer, H.G. 14
Gaiser, T. 120n
2.84, 2.94, 2.96, 359_60 Healy, D. 84
Cicourel, A 58, 81-2., 141
Dreyfus, H. 35 Heath, C. 6, 223, 240,
Clancy, S. 2.48 Gale, J. 295 260 ,
Duchon, D. 331 267, 272-3, 277-80,
Clarke, A 31 Garcia, A. 226
Duguid, P. 327, 329 287-8, 295, 323, 347, 359
Garfin kel, H. 35-6, 38,
Clarke, V. 2.05
Dura nti, A. 226 81, Hepburn, A 207
Clavarino, A 351 141 -2, 149, 155, 206,
Durkheim, E. 10, 15, Herbert, C. 181
Clayman, S. 209, 2.2.6, 2.38, 28n, 248, 268-9, 347, 350,
268, 286, 350-5 1 Heritage, J. 5-6, 35, 42,
240, 2.98-9 1 13,
Gajjala, R 1 1 8
Dwyer, K. 17 355, 358-9
Clifford, J. 14, 1 7 120n, 142, 188, 203, 209,

Eagleton, T. 36
Baker, C. 5 , 7, 156, 158, 163, Brown, J. 327, 329 Coffey, A 3-4, 3 1 , 358, 364 Geertz, C. 15, 20 222, 224-6, 229, 231 _3,

Eberle, T. 342
Bryman, A 362 Gergen, G. 284 235, 237-8, 240, 269_70
165, 1 68-9, 175, 342 Collins, H. 2.8n ,
Baldock, J. 229, 238 Button, G. 207 Collins, P. 131-2. Gergen, K. 104, 1 12, 272, 284, 286, 289, 291_
4,
Edwards, D. 202-3, 210, 202, 296, 298-3 00, 346, 353,
212,
Barbour, R. 178, 196
Banks, M. 262n Brooks, 1. 177 Comte, A 305 363n
217 359, 364n
Brown, G. 83, 201 Conley, J. 43-5, 48 Gidd ens, A. 104
Eglin, P. 207
Eliot, TS. 305
Herodotus, 12
Barnard, M . 262n, 322 Bruckman, A 1 1 8 Contreras, J. 17 Gilbert, N. 202, 211
Ellio tt, L. 1 77
Frerring, S. 1 04, 121
Bryman, A 195n Gilder, G. 102
Barrett, R. 50, 87 Gil l, R 205
Barthes, R. 77, 258, 262n Converse, J. 146-8, 150, 157
Emerson, R 40
Hes ter, S. 207
Heywoo d, 1. 262n
Bulbeck, C. 258 Cooper, P. 181

Hilbert, R. 36
Gill, V. 239-40
Burrows, R. 100
Bashir, M. 208-9, 215 Bulmer, M. 306 Corcuff, P. 30n
Emmison, M. 6, 25 1, 256
Engestrom, Y. 280
Glaser, B. 18, 23, 30n,
Buttny, R. 296
Baszanger, 1. 3, 19, 31n Cosnier, J. 267 286,
Esp in, O. 195n
Hindmarsh, ]. 267, 272-3
Bateson, G. 252 Coulon, A 40 330
Hine, C. 121
Glassner, B. 4, 129-30,
Coupland, N. 2.01
Baym, N. 99, 103, 120n, 121 Button, G. 35 Coulthard, M. 2.01
Evans, ]. 262n 133 Holstein, ]. 4-5, 40, 45
Becker, H. 30n, 251, 306, 320 Byrne, P. 229 Goffman, E. 20, 22, 30n, -6, 5 1,
Evans-Pritchard, E. 9 88, 140 -1, 143-4, 149--152,
Benedikt, M. 100 Cressey, D. 314 191, 222, 247, 252, 262,
Fairclo ugh, N. 285
Callon, M. 23, 30n 155-8, 1 62, 175, 1 77, 1 8
Berg, M. 78, 88 Cressey, P. 30n 267, 271, 286, 289, 350, 2,
188 , 343
Bergen, B. 43 Cambrosio, A 78-9 Cushman, D. 1 6 Farquhar, C. 180 352-4
Houtkoop-Steenstra,
Berger, P . 142 Campbell, A 134 Gonzales, P . 239 H. 82,
Favret-Sa ada, J. 17 152, 1 57, 296
Bernstein, B. 65 Carey, J. 305-6 Dalton, M . 326 Goo dy, ]. 92
Fea therstone, M. 100 Huberman, A. 360
Goo dwin, C. 223, 267,
Hud son, ]. 1 1 8

371
NAME INDEX

NAME I N D EX
Hume, M. 345 Latour, B. 23, 28n, 30n, 69, Mason, J. 248 238-40, 295--8, 300n,
Ryen, A. 95, 101, 1 75
Hutchby, 1. 210, 300n 88, 126 Mauss, M. 29n 322, 344, 359-60 Smith, P . 25 1, 254
Hyman, H. 140, 143 Lave, J. 325 Maynard, D. 41, 157, ] 66, Piette, A. 15, 22, 30 Soja, E. 99
n Sa cks, H. 7, 35, 40,
LeBaron, C. 267, 272 239, 269, 280 292-3, 295, 1'ilnick, A. 267 155, Sombart, W. 30n
Lee, ). 35, 42 359 Pink, S. 248, 262n 1 63-4, 189, 206-7
Inglis, K. 258 , 209, Sondheim, A. 100
lser, W. 89 Lee, M . 92n Mays, V. 18l Platt, J. 18, 20 2 1 1 , 223, 225, 228, 235,
Sorjonen, M.-L. 235,
Lehrer, A. 59 McEvoy, S. 22 Plummer, K. 92 268-70, 285, 289, 291, 284
Spanish, M. 183
Jackson, J. 50 Leiris, M. 28n McHoul, A. 45, 202, 226, 256 1'ollner, M. 84, 142 299, 342, 347-64
Speer, S. 205
P omerantz, A. 1 74 Sampson, R. 259, 2 62n
Lengua, L. 181
Jacobs, J. 37 Lempert, L. 130 McKeganey, �. 3] 2-3, 317, , 234, Spro ull, L. 1 20n
Sanders, C. 1 26
Mead, G. rI. 353-4, 358
Jacoby, S. 239 322 290--2, 295, 300n Stacey, R. 325
Sand strom, K. 128
Levi-Strauss, c. 13--15, 20, 77,
James, H. 342, 345, 347 Levinson, S. 211 Pool, I, 1 49, 152 Stake, R. 332
Postman, �. 1 1 9 Sandywell, B. 262n
j anowitz, M. 306 Mead, M. 9, 252
Potter, J. 5 , 1 89, 202-7
Star, S. 61, 78--9, 87
Saussure, F. de 363
79
Leydon, C. 7, 363
jarrett, R. 180 Mehan, H. 35, 226, 362 Stasz, C. 250
, Scha effer, �. 1 57
J aworski, A. 201 Meltzer, H. 81 2 1 0-1 1, 2 1 6 -1 7, 346 Stenning, P. 262
Powney, J. 181 Scheflen, A. 267
Jayyusi, L. 164 Licoppe, C. 1 1 Merry, S. 37, 44 Stewart, D. ] 80
Press, A. 195n Schegloff, E. 1 64, 166,
Lincoln, Y . 343
Jefferson, G. 42, ] 79, ] 89, 235, Lincoln, A . 1 1 9 Merton, R. ] 95n 189, ' Stewart, F. 99,
204, 21 1, 216, 222-4, 103, 1 1 8-19,

Lindstrom, A. 231, 286, 296

268-9, 289 Messinger, S. 40 Prince Charles 215 121
228, 235-6, 246, 268
Lindstrom, 1,. 35
Middleton, D. 280 Princess Diana 208-
Johnson, A. ] 78
jewi tt, C. 262n 10, 212, _9, Stimson, G. 255-6
Mi les, M. 360 215 289, 293-5, 298
Schiffrin, D. 201 Stiv ers, T. 229, 237, 240
j ohnson, C. 95, 102 Linhart, R. 29n Miller, D. 121 Prior, D. 229, 238
Pri or, 1,. 4, 58, 74, 76,
Stone, A. 1 00
Schmitt, D. 129, 134
Johnson, J. 144-5, 289 Linell, P. 236-7 Miller, G. 3, 6--7, 37, 40, 45--7, Straits, B. 1 44
84, Schon, D. 307, 321
Jones, S. 99, 12] Linstead, A. 344, 364n 49, 51, 296, 332 86-8, 92, 1 13, 261, 358
Prosser, J. 262n
Stra uss, A. 1 8-9, 22-3,
Schuman, H. 146-8, 30-- 1n,
Schu tz, A. 37, 152, 307
Jordan, M . 78 Lipsky, M. 326, 332-3 Miller, J. 4, 131-4, 1 36-7, 150, ] 57
Prout, A. 312
286, 330
Jorgenson, M. 201 Lishman, ) . 326 138n Streatfield, P. 325, 334,
Schwartz, D. 248, 258 337
Livingstone, S. ] 77 Miller, M. ] 45, 283, 289-90, Prus, R. 1 26, 129 Streeck, J. 267
Psathas, G. 204-5 Schwartz, H. 37
Kanter, R. 326, 330 Lofland, j . 31 294, 299 Strong, P. 240
Scott, J. 63, 74, 92
Kaplan, B. 331 Lofland, L. 31 Miller, �. 88-9 Puchta, C. 189 Stubbs, M . 202
Seale, C. 1 82, 300, 360
Katzman, S. 95, 120n Long, B. 229 Mills, C.W. 128 Sturken, M . 262n
. Sefi, S. 229, 23] , 238
Keating, 1'. 78 Loughlin, J. 1 29--30, 133 Mirzoeff, � . 262n Queen Elizabeth 215 Sudnow, D. 58, 286
Senft, T. 1 0 1
Kelling, G. 259 Luckmann, T. 142, 236-7 Mishler, E. 1 30, 141, 236-7, Swain, G. 1 77
Sharrock, W. 257
Kellner, D. 36 Luff, 1'. 223, 267, 272-3, 239, 332 Rabinow, P. 35
Shamdasani, P. 1 80
Rapley, T. 1 75, 364n
Lunt, 1'. 177
Keogh, J. 165 278-9, 287, 323, 359 Moern1an, M. 1 7, 364n Tayl or, A. 1 32
Sharf, B. 1 1 8
Rau denbush, S. 259,
Kendon, A . 267
Kendall, 1,. 99, 11 3-14, 121 Montgomery, M. 201 262n Te M older, H. 207
Sharrock, W . 207
Lynch, M. 205 Morgan, D. 88-9, 1 77-80, J�awls, A. 334 ten Have, P. 205,
Raymond, G. 239 Shaw, I. 321, 323, 326, 288-9, 300,
Kenny, A. 290 Lyons, R. 181, 187 182, 195n, 196 331, 362
Ker Muir, Jr, W. 334-7, 338 Morley, D . 258 Rees, C. 58 339
Terkel, S. 143, 148
R einharz, S. 142 Shearing, C. 262
Kershner, V. 1 77 Maccoby, E. 144 Morrice, J. 315
Renesch, J. 325
Thackeray, W. 89
Shields, R. 99
Kiesler, S. 120n Maccoby, �. 144 Morton, A. 215 Thevenot, L. 1 1 , 1 6, 19
R ichardson, J. 262n
Shumway, D. 35 , 22-3,
King, J. 181 MacDonald, J. 182, 187 Moss, 1'. 48 30n
Thomas, R 344, 364n
Siang, S. 1 1 8
King James 76, 91 Madge, J. 144 Mulkay, M. 202, 21 1 Richardson, L. 13 0
Ricoeur, P. 12, 14, Silverman, D . 1-2,
Kirk, ) . 145, 283, 285, 289-90, Maeder, C. 343 Murphy, E. 6-7 1 7, 2 1 19, 37, Thorne, B. 1 34
Maget, M. 29n Myers, G. 189 Riessman, C. 1 27, 130 42, 45-6, 59, 86, 97, ]
294, 299 , 1 56 03, Thurlow, C. 1 20n
Kissling, E. 180 Maguire, M. 58 Ritvo, H. 78 120, 1 25-7, 138, 140_1,
Thrasher, F. 29n
Malinowski, B. 9, 1 6 luviere, P . 39 156-8, 1 63, 166, 1 68-9
Kitzinger, C . 188 N avel, G. 29n , Thurstone, W. 1 95n
Kitzinger, J. ] 78-9, 181, 184, Mangione-Smith, R. 240 �ewman, D. 246 Roberts, G . 58 182, 206, 23 1, 239, 283,
Tuck, M. 194
Mangone, T. 143 �ovak, M. 99 Robin son, D. 262n 285, 289, 292, 296_3
1 87, 196 00,
Robinson, J. 229, 237,
TUrkle, S. 99-1 00, 121
Mann, C. 99, 103, 1 18--19, 121 306-7, 322-3, 344, 347,
Kitsuse, J. 58 240
Rock, 1'. 31 8 359-60
Oakley, A. 1 8 1
Kleinman, S. 130 Manning, P. 141, 318 O'Barr, W. 43-5, 48 Unruh, D. 38
Ussher, J. 195 n
Simmel, G . 30n,
Knorr-Cetina, K. 28n Marcus, G. 16-17 Rogers, C. 144 260-1, 352
Rogers, R. 128 Sinclair, J. 201-2
Knott, B. 1 77 Marks, D. 271 Orr, J. 327--8, 338
Singleton, R. 144
Kroeber, A. 262n Markham, A. 4, 98, 1 1 5-] 6, Rogers, W. 128 van Dijk, T. 201-2
Slater, D. 121
Kroger, R. 201, 207 121 Parker, 1. 285 Rose, E. 1 13 Vaughan, D. 326, 333,
Ros e, � . 84 Small, A. 250 339
Krueger, R. 179-81, ] 96 Marlaire, C. 333-5 Parsons, T. 358 Vehviliiinen, S. 296,
Smith, D. 38-9, 48, 300n
Kunda, G. 337 Marshall, H. 205 Patton, M. 326, 330 Roy, D. 326, 338 129, 142, Volkmann, R. 3 1 5
Marx, K. 286 Perakyla, A. 6, 42, 204, 227, Rushdie, S. 2 1 1 287
Kvale, S. 158 vom Lehn, D. 2 60,
Smith, M . 183-4 267
Vurd abakis, T. 58
372

373
NAM E IND EX
210 , 300n
W ooffitt, R. 206-7,
Whyte, W.F . 28n Woolgar, S. 28n, 69
Wagner, J . 251 Widdicombe, S. 206
-7
R. 258 Wootton, A. 300n
Wagner-Pa cifici, Wiener, C. 328-30
Young, A 31n, 84,
144
Wats on, R. 1 13, 238 , 2
Warren, C. Wilkinson, S. 5-6, 177
-9,
87
49, 296
182, 184, 195n , 196
195 n Yule , G . 20 1
Webb, E. 259
W eaver, J. W illig, C . 207
Wils on, J. 259 Zeller, R. 294
Weber, M. 356
Witt genstein, L. 290-
Witmer, D . 95 Zerub avel, E. 87
Weil , S. 29n 1,
Ziller, R. 248
Weitz, R . 131 363 42, 84,
Wenger, E. 325-6 Zimmerm an , D . 35, accountants 3, 57, 66·-7 Chicago School 18-19, 29n, counselling 46, 144, 173,
Wong, G . 1 8 1 203, 239 , 269 , 299 , 34 5-7,
West, c. 299 accounts 27, 125, 163, 1 74, 305, 352, 354 296-7, 309, 344
W ood, H . 35
Wood, L. 201, 207 Zola, 1. 237
7, 216 356 215, 364n childhood, children 128, 240, crime 58, 81, 137
W etherell, M. 201-
W halen, J. 253, 286, 31 3, 326, 333�4,
237 , 238-9, 272-3, adjacency-pairs 293, 296 culture, cultural 3, 13, 15·-18,
Woods, P. 58
287 adolescents 1 28-9, 131-8, 364 30n, 37, 44, 47, 59, 61,
1 63, 1 68-9, 313 cities 23 65, 72, 74, 77, 96-1 01,
advertisements 250, 255-6, claims 2 1 1 1 13, 1 17, 1 19, 1 30, 1 34·-5,
258 class 48, 50, 127-8, 247-8, 1 37-8, 150, 163, 1 6 7, 169,
AIDS 46, 89, 131, 1 77, 1 79, 253-4 1 75, 246, 248-9, 252, 254,
187, 297, 309, 31 1-12 classification 61, 78, 80, 86, 256-8, 262, 296, 332, 344,
ambulatory events 286-7 241 347, 358, 364n
analytic induction 362 classrooms 203, 222, 226, 284 curricula vitae 89-90
anorexia 95, 102 codes 249 cyberspace 95
anthropology 9, 1 1 , 13-14, coding 10, 31n, 1 27, 1 83,
24, 56, 78, 140, 246, 272, 204-5, 207, 216, 258, 298, delinquency 29n, 133
354, 364n 355 dementia 152
architecture 99, 248, 261 cognition, cognitive 207, 210, deviance 128
attitudes 354 216, 269, 280 deviant cases 204-5, 284,
audit 60-72 commentator machine 292-3, 299, 359
authentic(ity) 1 02-3, 125-6, 349-51 diagnosis 85-6
138, 146, 155, 1 62 communication 42, 95-124, dialogue 1'7, 38, 72
author, authorship 17, 70-3 144, 149, 177, 1 80, 202 diaries 1 7
comparison 1 3, 1 8 , 46-7, 313 discourse(s) 25, 36-7, 39,
behaviourism 347 concepts, conceptualization 42-3, 47, 49, 83, 99,
beliefs 361 83, 205, 255 120n, 126, 184, 202
bias 157 confidentiality 131, 134, 1 79 discourse analysis 5, 35, 156,
billboards 255 conflict 46 179, 186, 195n, 200-1 8,
biographies, biographic, consumption 73 285
biography 72, 1 30, 1 49, content analysis 5, 1 79, documents 56-74, 76-92, 1 18,
179, 195n 1 82-7, 189, 1 94, 1 95n, 203, 284, 287, 339
body, the 100, 222, 266-7, 205 drugs 181-2, 313
270-71, 274-7 context(s) 41-2, 72, 99, 1 12,
boundaries 79, 84, 100 1 1 9, 129, 187, 216, 223-6, ecology 18, 46, 250
blaming 210 269-70, 294-5, 300, 332, economy, the 256
breast cancer 1 84-194 334, 352, 360 education 3, 56, 58, 65, 132,
bureaucracy, bureaucratic conversation analysis 5, 1 65-7, 1 75, 222, 227-36,
60, 68, 72, 326 12-13, 35-7, 40-2, 44-5, 333-5
business 287, 360 47-5 1 , 1 20n, 1 79, 1 88-9, empathy 14-15
200, 202-10, 216-7, emotionalists 125
cartoons 249, 256 222-241, 260, 262, 266, emotions 15, 1 46-8, 150
case studies 3, 1 14, 296 268-73, 278-9, 284-300, encyclopaedias 77-80
categorization 39, 84, 237, 333, 339, 359-60 environment(s) 257, 261, 266
294-5, 298-9, 349-63 corporate governance 333, epidemiology 318
caution 234, 238-9 337 epistemology 1 1 , 23, 27, 119,

374
 S U BJECT /NDEl(

SUBJ ECT IND EX

140 66-8, 73, 87, 1 03, 203, readership 70-73 society 16-17, 20, 56, 354
life history 29-30n,
29n, 1 0 1-2, 1 13, 1 1 8, 300n, 325-39, 342-4, recordings 38, 42, 44-6, 59, sociology 3, 9, 16, 18-19, 26,
1 8 7, 1 94, 195n, 207,
238, linguistic registers 59 360, 364n
120, 248, 261 349, 133, 165, 1 79, 251, 272, 28n, 30n, 36-7, 51, 74,
256 linguistics 201 -3, 206, otherness 26-7
gestures 270-1 283-300, 353--6 86, 140, 156, 200, 203,
etcetera principle 350-1 355
records, record-keeping 57, 223, 247-8, 251, 261 , 266,
glances 261 3-4,
ethical, ethics 1 1 8,
180, 325, literary theory, criticism paradigm(s) 51
grammar 201, 203 68, 74
61 , 67-9, 86-8 268--9, 280, 305-6, 314,
338-9 pastoral care 3 1 7--1 8 reductionism 26
greetings 205 317, 321, 339, 350-2, 354,
ETHNOGRAPH 1 79 330 local cultures 40
grounded theory 23, 1 03, 286 performance 50 reflexive, reflexivity ] 7, 36, 356--7
ethnography 3, 9-51,
56, 59, longitudinal studies pharmacists 89-90, 343 41, 1 1 2, 1 1 4, 130, 202, space 18, 26-8, 29n, 99, 26]
1 14, 1 26, 178, 1 82-3,
Harlemworld 50 330, phenomenological, 250, 252, 270, 280 stakeholders 7, 326, 328, 333,
187_94, 195n, 246, 248
,
3, 28 6 mana gement 7, 325-6,
92, 299, 305, health visitor(s) 232- 332-4
phenomenology 50, 1 '19 reliability 6, 204, 283-9, 337, 339
272, 2 85-6, 2 20, 36 philosophy 1 1 , 36, 290, 339,
hermen eutics 14, 299-300 stake inoculation 212, 215
308-23 , 339, 342-7, 29n, 44, 59, maps 249, 256-7
history 17-18, 23, 350 repair 38 statistics 16, 26, 28n, 30n,
, marketing 177
349-50, 352, 354, 360 1 17, 203, 254, 256, 259, photographs 246-55, 262 representation(s) 1 1 , 63, 65, 5'1-8, 81, 84, l O2, 183,
364 market rese arch 1 77 police 224, 234, 238, 259,
286 141-2, 80, 141, 156, 249, 25�, 259, 298-9, 362
ethnometho dology
9, 1 1 , 1 3, 321 , meanin g(s) 4, 76, 1 13,
hospitals 84-8, 315-6, 6, 223 , 29 1 , 344, 335--6, 340, 352-3 363n stories 127, 130, 135--8,
22, 27, 30n, 35-40, 42, 325, 328-9, 33 1-2, 339
145, 1 5
policy-makers 305-7, 31 8--22, rhetoric 16, 40, '11, 73, 202, 190-], 202-3, 343, 349,
-9,
44-51, 141-2, 163, 188 raction 364n
human-computer inte 2 326 207, 359, 363n 357--8
206_7 210, 217, 249, m easurement 24-7, 8 politics 132
279 romanticism ] 27, 1 38, 1 62, subculture 3, 13, 37, '12
256_7, 260, 262, 266, 216 media studies 126, 1'17
humour 29n, 1 1 5, 188, 29n, 3 1 n, positivism, positivist 1 25--6, 344--5, 360, 36311, 364 suicide 81, 268, 315
268_9, 272-3, 278-
9, 339, medicine 6, 1 9 , 21 , 1 82
4, surveillance 3] 8
40, 43, 58, '12, 8'1, 193-
-;;
344, 358, 363n 03, 11 5,
126 , identity 21, 77, 1 00, 1 201, 212, 222, 234, 238-
9, poststructuralism 36, 141 , sampling 23, 298 system(s) 78
b1,
experience(s) 5, 1 14, 129, 153, 205, 248--9 201, 253, 344 schools 49, 58
144, 146, 25'1, 269, 2 '13-8, 284,
129, 138, postmodernism 36, 141, 307,
incongruity 353 286-8, 298, 30'1
science 9, 1 1-12, 2 1 , 23, 27, technicians 327-8, 331
155, 342-64
indexicality 269-70 ation 363n 28n, 69, 80, 30'1, 309, 3 1 4 technology 78, 80, 97, 99-100,
experiment(s) 205-6,
209 membership cate goriz
information 78 86, power 43, 46, 48-9, 181, 241, script(s) 86, 201 ] 02, 104-5, 1 1 0, 1 13, 1 1 '1,
ulation 174 analysis, devices 5,
extreme case form institutional discourse
/talk 318 self 21, 77, 88, lO6, lO8-lO, 1 1 9-20, 143, 2 1 7, 248,
1 62-'15, 206-'1
37 _8,42 -3, 45-51, 188, professions 57-8 128 273, 2'19, 287, 323, 337,
64, 178, memory 103
families 5, 86, 1 52-4, 1 2 10, 222-41 , 284, 29
4-9 prostitutes 7, 304-1 1 , 318-19 sequence organization 230--1, 346
45, 48,
181, 222, 232-4, 253, 3
13,
1 1 , 1 8-19, mental i llness 38-9,
interactionism 9, 50, 8 0-4, 26 1 , 360
proxemics 262 241, 277 telephone calls 206, 227-236,
335 ,
30n, 126, 129, 131, 269 psychiatry 40, 50, 82-8, 140, semiotic(s) 59 272, 287, 293, 298, 345-6,
334-5, 360 metaphors 205
family ther apy 46, 343 211, 313-14, 344 sexuality 102, 131, 1 34-5, 181, 348-9, 356
fault(s) 231 124 micropolitics 40, 48
internet 4, 95- psychoanalysis 17, 150, 296 204, 307-12, 315 television 30n, 119, 1 34-5,
m 48, 142, 177, 344 184, motive (s) 210
feminis interpretive repertoir
es psychological, psychology skill(s) 22-7, 333, 336 140, 203, 258, 344--5
260-1
fiction 1 6, 89 21 1 museums 248, 252-3,
205, 207, 40, 140, 169, 187, 195n, social construction(s), social temporality, time ] 7, 22-3,
29n, 2 51,
fieldnotes 9-1 0, 16,
intersubjectivity 224 1 , 200-3, 206, 209-10, constructionism 5 1 , 1 14, 69, 103, 286, 338
283, 285, 313 narrative(s) 3, 1 0, '1-1 8
13-1 5, 17, 2 1 , intertextuality 66-9, 74 20_ 21, 24, 29n, 49, 131 ,
21 6-7, 278, 280, 285 120, ] 25, 141, 1 88, 191, text(s) 24, 44-6, 65, 68, 72,
field work (e r) 9, 44, 48,
309, interviews 4-5, 29n, 134, 141, 152, 1 56, 179,
psychotherapy 296, 298 202, 200 217, 285, 291 73-4, '16, 80, 97, lO4, 109,
26, 46, 56, 273, 279, 59, 82-3, 97, 1 02-4 , 106
, public health 3 1 8 social facts 10, 28n 1 1 6-18, 121 11, 127, 203,
316 201 , 249, 258, 364
1 14-1 6, 1 25-75, 178 , data 3, social problems 2, '1, 305-23, 205, 207, 212, 2 1 6, 26211,
film 272 naturally occurring
195n, 203, 205-7, 248, , 200 , 20 5-6, 262, quality issues 360-3 358-60 283-4, 299
fishing devices 290-
3 349, 138, 188
3, 105, 251 , 330-1 , 335, 339, quantitative data, methods social psychology 3, 1 '15, 200, theory 20, 25, 45, 51, 83, 126,
focus groupS 5, 102- 289
352, 360-1 , 3 63n, 364 30n, 83, 1 '18, 184, 1 95n, 202-3, 206, 209, 2'18 182, 186, 189, 1 91 , 1 93,
1 1 0, 1 1 4, 1 77-196, 248
, nat ural sciences 1 1
ips eity 21 204, 240-1, 259, 262n, social science(s) 1-2, 12, 1 7, 200, 205, 249, 254, 256,
360 navi gation 257
298-9, 326-'1, 330-1, 362 76, 90-1, 162, 168, 1 7'1, 259, 262, 278, 296, 325,
footin g 191 neuroses 80-4
journalism 298 , 337, 357 6, 238, questionnaire(s) 80-] , 205, 182, 204, 206, 248-9, 339
framers) 30n , 36 news interviews 225-
justice 23 251, 289 252, 261, 266, 268, therapeutic communities 7,
269, 29 6
questions 230-40, 29'1-8, 325-6, 332, 334, 336, 347,
time see temporality
313-19
gangs 1 32-5, 137 205, 331 , 352 newspapers 215, 35 7-8
laboratory 79, 347-8 364n
gaze 222, 262, 3 1 7 59, norms 204-5, 223, 293
, 1 4, language, 1 2, 42, 47, 49, ST 1'19
socialization 20, 29n totality, totalization 1 0, 13,
gender 48, 99-1 00 1 NUD.l
, 259, 1 08, 120n, 1 62-3, 175, 86, 150-1
race 48, 50, 127, 131-2 social structure 203 16, 19
127_8, 131, 134, 249 203, 266, 285, 29 1 , 29 6-7, nursing
radio 11 9, 296 social surveys 146-8, 15'1, transcription, transcripts 6,
299
335-6, 347-63 , 363n rape 95, 99 251, 330 42, 1'19, 183, 203--4, 207,
generalization 18-1
9, 206, objectivity 282
lawcourts 4 1 , 43, 45, rapport 132-4, 1 62 social theory 2 2 1 0, 216-17, 269, 273,
295 -300 opera 91-2
225-6, 238 61 , 63, rationalization 61, 63 social worlds 38 284-300, 361
genre 3, 59, 249 239, 241 organization(aD 56-9,
grap hic 1 8, lexical choice 235-6,
geograp hy, geo
377

376
SUBJECT INDEX

translation 80
triangulation 35-6
turn design 231-4, 240
turn-taking 225-6, 241, 270,
290-3, 359
underdogs 49
universities 60-72
validity 6, 15, 59, 141, 145,
207, 283-5, 289-300, 349,
362
variables 204
video cameras 101
video(s) 104, 209, 260,
266-80, 286-8, 317
visual research 6, 246-80
web sites 57, 101, 1 l 0-11
wine 59
work 31 n, 49, 143
workplace studies 223, 273,
287, 296, 323

378