vl

Like all academic work, this book only exists because of the help and
creative insights of a number of other people. we are particularly indebted
to fwo friends and colleagues with whom many of the ideas that appear
in
this book were first developed, Tom Lawrence and steve Maguire have
been our coauthors in a stimulating and productive collaboration that
formed the foundation ofour understanding ofdiscourse analysis. over
has
the
DISCOURSE ANALYSIS
course of several research projects, aird a number of books and articles,
we
have learned to apply and explain discourse analysis by working with
Thanks, guys-we owe you a beer.
them. Investigating Processes of Social
we would also like to acknowledge Davide Ravasi, Ian palmer, and Construction
stewart clegg' with whom we have written papers we refer to in this book,
as well as Susan Ainsworth, Andr6 Spicer, David Grant,
and cliff oswick,
who reviewed an early draft of the manuscript. Reading a first draft I
is never
an easy task, and we appreciate the time and effort you all put
and commenting on this book. we would also like to thank the
into reading
editors of
NELSON PHILLPS sqiP P"blic"\;on,
sage's Qualitative Research Methods series, especially John van Maanen, Univ ersity of Cambridge
for their encouragement and insight. on a pragmatic note, this book was 2oo>
written while Nelson phillips was a Dyason Universitas 2l visiting Fellow CYNTHIA HARDY
at the university of Melbourne. we would like to acknowledge
the university Univ er s ity of M elb ourn e
of Melbourne's support, which made the writing of this book possible.
Finally, on a more personal note, Nelson would like to thank Deana
for her
consistent good humor and support despite missed dinners and curtailed
vacations' He promises not to write another one for a while. cynthia
would
like to thank Jerry for missing that bus-who knows what would
happened to the book otherwise.
have
I. WHAT IS DISCOURSE ANALYSIS?

Her knowledge of me was so deep, her version so compelling,

that it held together my miscellany of identities. Tb be sene, we
choose between the diver.se warring descriptions of our selves;
I chose hers. I took the name she gave me, and the criticism,
and the love, and I called the discourse me.

Salman Rushdie, The Ground Beneath Her Feet(2000, p. 510)

This book is about discourse. More specifically,it is about the power of

incomplete, ambiguous, and contradictory discourses to produce a social
2 3

reality that we experience as solid and real. we understand discourse in the diftlculties of applying discourse analysis to different contexts. By writing
radical, constitutive way of Rushdie's character: The things that make up this book, we hope to save other researchers from having to go through the
the social world-including our very identities-appear out of discourse, same struggles. By providing a general framework for understanding differ-
To put it another way, our talk, and what we are, are one and the same. But ent forms of discourse analysis and applying them to empirical studies of
we differ from Rushdie's character in that we do not believe individuals organizational, interorganizational, and societal phenomena, we hope to
always have the luxury of choosing their identity, their truth, and their real- save other researchers from having to "reinvent the wheel."
ity. We think our experience is largely written for us by the multitude of
conflicting discourses of which we are a part. This is not to say that we do
not strategically draw upon these discourses. We obviously do. But our Defi ning Discourse Analysis
ability to act strategically is limited by the discourses that accompany our
intervention and the complex processes of social construction that precede There are many definitions of discourse and discourse analysis in the
it. our view of discourse can be summarized in a sentence: without dis- literatute. In fact, in his introduction, van Dijk suggested that the entire
course, there is no social reality, and without understanding discourse, we 700 pages of the recent two-volume set on discourse (1997a, 1997b) is

cannot understand our reality, our experiences, or ourselves.

really an "elaborate answer" to a deceptively simple question: What is dis-
course? Yet, despite the diffrculty of the task, we need sorne general idea of
Recognizing the profound role of talk and texts in everyday life is only the
beginning. This book is also about the process of analyzing discourse and what we are referring to when we use discourse analysis and related terms.
We also need to differentiate between discourse analysis and other qualitative
the potential of this methodology for revealing the processes of social con-
struction that constitute social and organizational life. Discourse analysis methods that explain the meaning of social phenomena. In this section, we

offers new opportunities for researchers to explore the empirical ramifi- present some of the important terms that relate to discourse analysis. We
also describe its status as a methodology rather than just a method, that is,
cations of the linguistic turn that has worked its way through the social
an epistemology that explains how we know the social world, as well as a
sciences and humanities in the last 20 years. whereas other qualitative meth-
set of methods for studying it. In this way, we differentiate discourse analysis
ods provide well-developed approaches for understanding thc social world
and the meaning it has for the people in it, discourse analysis goes one step
from other qualitative research methods, such as ethnography (Erickson &
further in embracing a strong social constructivist epistemology (Berger & Stull, 1997; Schwartzman, 1993), ethnomethodology (Coulon, 1995), con-
Luckmann, 1967; Gergen, 1999). It focuses attention on the processes versation analysis (Psathas, 1995), and narrative analysis (Czarniawska,
1998; Riessman, 1993).
whereby the social world is constructed and maintained. It also includes the
academic project itself within its analysis; with its emphasis on reflexivity,
discourse analysis aims to remind readers that in using language, producing Defining OurTerms
texts, and drawing on discourses, researchers and the research community Discourse, in general terms, refers to actual practices of talking and wrir
are part and parcel of the constructive effects of discourse. ing (Woodilla, 1998). Our use of the term is somewhat more specific: We
We wrote this book for three reasons. First and foremost, we find dis- define a discourse as an interrelated set of texts, and the practices of their
course analysis to be a compelling theoretical frame for observing social production, dissemination, and reception, that brings an object into being
reality. This book represents our attempt to clarify the contribution that dis- iTindi that
course analysis can make to the study of individuals, organizations, and .rnake up the discourse of psychiatry brought the idea of an unconscious into
societies. second, we have found discourse analysis to be a useful method existence in the l9th century (Foucault, 1965). In other words, social reality
in a number of empirical studies and have increasingly adopted it over the is produced and made real through discourses, and social interactions can-
-
last l0 years. we want to encourage other researchers to adopt this approach
and believe that a short, simple introduction will help in this regard. Third,
@"rer"nie tnTtJ discourses thai give them

we have spent considerable time over the last l0 years struggling with the b
4 )

Discourses are embodied and enacted in a variety of texts, although they discourses have evolved over time, we can study texts such as cartoons,
exist beyond the individual texts that compose them. Texts can thus be newspaper articles, and international conventions. We must also examine
considered a discursive "unit" and a material manifestation of discourse the social context-wars, natural disaster, court decisions, international
(Chalaby, 1996). Texts may take a variety of forms, including written texts, agreements, the govemment of the day, political events in other countries-
spoken words, pictures, symbols, artifacts, and so forth (Grant, Keenoy, & to see how they are brought into play in particular discursive events. This
Oswick, 1998). interplay between text, discourse, and context helps us to understand not
only how an individual comes to be a refugee, but also how the broader
Texts are the sites of the emergence of complexes of social meanings, produced ..reality" of refugee policy and refugee determination procedures is
in the particular history of the situation of production, that record in partial
ways the histories of both the participants in the production of the text and of
constructed and experienced.
the institutions that are "invoked" or brought into play, indeed a partial history In summary, our interest in the relation between discourse and social
of the language and the social system, a partiality due to the structurings of reality requires us to study individual texts for clues to the nature ofthe dis-
relations of power of the participants. (Kress, 1995, p. I 22) because we can never find discourses in their entirety. We must
"our("
therefore examine selections of the texts that embody and produce them
Ig$g_"I9!.ggg"stuIindtftd"ully;it @geLSgI!nt9r,999nry!og (Parker, 1992). We cannot simply focus on an individual text, however;
with other texts, the different discourses on which they draw, and the ni4ure rather, we must refer to bodies of texts because it is the interrelations
of me"ptoauction, Aiss"*iration, anO cr at tfrey -" rnuO*"-- between texts, changes in texts, new textual forms, and new systems of dis-
ingful. Discourse analysis explores how texts are made meaningful through tributing texts that constitute a discourse over time. Similarly, we must also
these processes and also how they contribute to the constitution of social real- make reference to the social context in which the texts are found and the
iryby making meaning (Phillips & Brown, 1993). discourses are produced. It is this connection between discourses and the
Discourse analysis is thus interested in ascertaining the constructive social reality that they constitute that makes discourse analysis a powerful
effects of discourse through the structured and systematic study of texts method for studying social phenomena.
(Hardy, 2001). Discursive activity does not occur in a vacuum, however, and
discourses do not "possess" meaning. Instead, discourses are shared and
social, emanating out of interactions between social groups and the complex
Discourse Analysis as Method and Methodology
societal structures in which the discourse is embedded. Accordingly, if we The reason discourse analysis tries to include a concem with text, dis-
are to understand discourses and their effects, we must also understand the course, and context relates to the fact that it represents a methodology-not
context in which they arise (Sherzer, 1987; van Dijk, 1997a). just a method-that embodies a "strong" social constructivist view of the
social world (Gergen, 1999). Discourse analytic approaches share an inter-
Discourse is not produced without context and cannot be understood without
est in the constructive effects of language and are a reflexiv€--as well as an
taking context into consideration. . . . Discourses are always connected to other
discourses which were produced earlier, as well as those which are produced interpretive-style of analysis (Parker & Burman, 1993). In this regard, dis-
synchronically and subsequently. (Fairclough & Wodak, 1997 ,p.277) course analysis does not simply comprise a set of techniques for conducting
strucrured, qualitative investigations of texts; it also involves a set of
Our approach to the studl of discourse is therefore_'$:qg:gryglqio!-al" assumptions concerning the constructive effects of language.
1992), in $g_.seqse-{4-i1 cgtngqts_ggxts_t_q d!_sgous_e_q, l_qc_elitg
'€qttqlgugh,
them in a historical and social context, U t!g!_ry"-Ie.!er1g theg4pUlar [Discourse analysis] is not only about method; it is also a Perspective on the
illoisJelationsEps, and practices that char4qJe-rlz-e-ttre-situjtionJndelstrrdy-. nature of language and its relationship to the central issues of the social
sciences. More specifically, we see discourse analysis as a related collection
Consider an example: To understand from a discourse analytic perspective
of approaches to discourse, approaches that entail not only practices of data
why a particular person is a refugee, we need to explore how discourses collection and analysis, but also a set of metatheoretical and theoretical
such as asylum, immigration, humanitarianism, and sovereignty, among assumptions and a body of research claims and studies. (Wood & Kroger,
others, serve to make sense of the concept of a refugee. To learn how such 2000, p. x)
 6

Discourse analysis shares the concem of ail qualitative approaches

with
the meaningfulness of sociar rife(winch, l95g), but it attempts to provide a Box 1.1
more profound intenogation of the precarious status of meaning. Traditional
Example: The Analysis of Globalization
qualitative approaches often assume a social world and then seek
to under-
stand the meaning of this world for participants. Discourse anarysis,
on the A quantitative study of globalization might involve collecting infor-
other hand, tries to explore how the socially produced ideas and
objects that mation on the degree to which globalization activities are evident
populate the world were created in the f,rrst prace and how they
are main- in a particular setting. Researchers might collect statistics on foreign
tained and held in place over time. whereas other quaritative methodologies
direct investment, the number of strategic alliances with overseas
work to understand or interpret social reality as it exists, discourse analysis
companies, decisions of the World Trade Organization, the use of
endeavors to uncover the way in which it is produced. This
is the most technology in developing countries, the size and nature of trade
important contribution of discourse anarysis: It examines how language
flows, or indicators concerning the prevalence of a global pop culture.
constructs phenomena, not how it reflects and reveals it. In other
words,
/. Such studies would attempt to connect the degree of globalization, as
discourse analysis views discourse as constitutive of the social
world-not denoted by these quantitative indicators, with particular outcome
a route to it-and assumes that the worrd cannot be known separatery from measures such as profitability, poverty, demographic trends, and so
discourse.
forth. Such research takes the concept of globalization for granted
Discourseanalysis is thus distinguished by its commitment to a strong and seeks to ascertain relationships among particular practices and
social constructivist view and in the way it tries to explore the relationships
outcomes to draw conclusions about the prevalence or effectiveness
between text, discourse, and context. Although studies vary
in the degree to of globalization.
which they combine text and context (as we discuss in chapter
z), ais- Qualitative studies of globalization can take a number of forms.
course analysis presupposes that it is impossible to strip discourse
from For example, an ethnography might involve a researcher living in a
its broader context and uses different techniques to analyze texts for
clues small village in a developing country to ascertain the meaning and
to the discourses within which they are embedded. In this regard,
discourse impact ofnew Internet connections for villagers; how the presence of
analysis is different from other forms of qualitative research. For example,
multinational companies affects family life; or how global calls for
approaches such as narrative analysis and conversational analysis
typically bans on child labor influence economic and social well-being. A
study text or talk, They take context into account to ascertain meaning,
but researcher could also undertake an ethnographic study of an Indian-
usually without reference to broader discourses or the accumulated
bodies based call center, in which he or she observes how employees present
of texts that constitute them. Although interested in how narratives
and con- themselves-via telephone-to callers from around the world and
versations are constructed, these approaches devote less explicit
attention what this means for the employees of the organization. One could use
to the construction of a broader social reality, similarly, ethnographies narrative analysis to uncover the stories that people tell to explain
often aim at uncovering the meaning of a sociar rearity for participants
but new global practices, using devices such as plot, narrator, and charac-
are less concerned with how that social reality came into
existence through ters to ascertain how they make sense ofa new Internet cAf6, an inter-
the constructive effects of various discourses and associated texts,
Ethno- national merger, or the sudden disappearance of an overseas market.
methodology focuses on the generative rules that make social interrelation_
One might use conversation analysis to sfudy teenagers in different
ships possible, but its focus is on the observation of actions rather
than on countries talking among themselves about what MTV means to them
the study of texts. In Box l.r, we provide an example to show
the ways in and what they consider important about their dress style. One could
which quantitative and qualitative researchers might approach phenomlnon
a conduct interviews with key actors in the World Trade Organization
and contrast them with how discourse analysts would study
it. or the United Nations to ascertain their views on the North-South
 Table 1.1 Diversity in Data and Traditions of Discourse Analysis
Box l.l (continued)
Examples of Data Exanples of Traditions
divide and compare their comments with the opinions of offrcials in in Discourse Analysis in Discourse Analysis
governments of southem counties. A
researcher's political analysis r r
Interviews Conversationanalysis
might use unstructured interviews and participant observation to r r
Focus groups Foucauldian research
uncover the politics and cultural dislocation involved in a takeover of r Documents and records . Critical discourse analysis
a local firm by a U.S. multinational corporation and to highlight any r Naturallyoccurringconversations r Critical linguistics
actions---covert or overt-by unions, employees, and community r Political speeches 'o Discursivepsychology
members to resist or influence the changes. Such qualitative studies o Newspaper articles Bakhtinian research
r Cartoons r Interacti<lnaliinguistics
are all, in different ways, interested in the social and political dynamics
o Novels r EthnoBraphyofspeaking
associated with globalization practices and in what those practices
mean to individuals who are affected by them. These researchers are SOdnCe: Adapted from Wetherell, M. (2001). Debates in discourse research. In
interested in the meaning, rather than the .,facts," of globalization, but M. Wetherell, S. Taylor, and S. J. Yates (Eds.), Discourse theory and practice:
.,given." A reader (p.38). Thousand Oaks, CA: Sage.
they still take the concept of globalization as
Discourse analysis is interested in how the concept of globaliza_
tion came about-why it has a particular meaning today when, It is important to note, however,that some traditional qualitative
60 years ago, it had none. Researchers might explore how globaliza_ approaches do lend themselves to discourse analysis. For example,
tion discourse draws from and influences other discourses-such as conversation analysis and narrative analysis can be used to connect
free trade discourse and liberalism, discourse around new technology, "microevents" to broader discourses as a way to show how narratives and
poverty and democracy, and even health and terrorism_and how it conversations construct social experience (e.g., O'Connor, 1995; Stokoe,
is constructed through diverse texts that range from academic articles 1998; van Dijk, 1993). Similarly, ethnographies have been an important
to CNN newscasts. They might then investigate how this broad dis_ component of discourse analytic studies in showing how discourses are
course ofglobalization gives meaning and substance to disjointed and enacted in particular practices (e.g., Covaleski, Dirsmith, Heian, & Sajay,
contradictory patterns of economic, social, geographic, and cultural 1998; Fletcher, 1998 Jackall, 1988; Orr, 1996). Content analysis, not in
activities. At the local level, researchers might explore how the dis- terms of a mechanistic counting but in a more interpretive form, can
course of globalization makes certain practices possible or inevitable- be used to connect textual content to broader discursive contexts. For
such as the business operations of multinational companies, restrictions example, Ellingson (1995) carried out a content analysis of newspaper
on refugees, or trade patterns between countries-and how it empow_ articles and editorials by identifying themes and rhetorical strategies and
ers and disempowers different identities. They may also investigate connecting them to the speaker and the audience; Holmes (1998) con-
how particular actors draw on the globalization discourse to legitimate ducted a content analysis of women's language use by linking it to power
their positions and actions. In exploring different texts pertaining to and status. Although the philosophy underpinning discourse analysis
globalization and relating them to the broader economic, social, and differentiates it from other forms of analysis, when it comes to actual stud-
political context, as well as to more specific practices, discourse ana- ies, the boundaries between discourse analysis and other qualitative meth-
lysts are able to draw conclusions that undermine the very notion of ods are sometimes blurred. Researchers have consequently employed a
globalization, showing how it is neither inevitable nor complete but, range of interpretive techniques-from microanalyses of individual utter-
in fact, a confluence ofdiscourses, texts, and practices that make up a ances to macroanalyses of a corpus of texts-to undertake discourse analy-
particular reality. sis and, as Table 1.1 shows, they have borrowed from traditional qualitative
methods to do so.
l0 lt

What makes a research technique discursive is not the method itself but location of individual texts in larger bodies of texts, and pay some attention
the use of that method to carry out an interpretive analysis of some form of to three-dimensionality. We are also faced with the prospects of learning
text with a view to providing an understanding of discourse and its role in by doing as we employ a particular analytical technique. interpreting mean-
constituting social reality. To the extent that they are used within a dis- ings as we go along and giving voice to multiple meanings. And, having
course analytic ontology and epistemology, many qualitative techniques incorporated all this into our study, we have to explain our work within the
can become discourse methods. confines of the normal avenues and arenas of academic publication. It is this
One final characteristic of discourse analysis is also worth noting: complexity and ambiguity that makes discourse analysis such a challenge-
Discourse analytic methods are unavoidably reflexive because the strong and also one of the reasons why we wrote this book, although by this stage,
social constructivist epistemology that forms its foundation applies equally the reader may be left thinking, why bother with discourse analysis at all?
to the work of academic researchers. Academic discourse also constitutes a
particular reality, and we are continuously challenged to retain a sensitivity
Rgasons for Using Discourse Analysis
to our role in the constitution of categories and frames that produce a reality
of a particular sort (see Marcus, 1994). Whereas other approaches tend to In this section, we discuss some of the reasons for using discourse analysis.
take analytic categories for granted and allocate data to them, discourse Given the plethora of other more established methodologies and the diffi-
analysts are interested in the socially constructed nature of the research culties noted above in doing discourse analysis, why should anyone con-
categories themselves. sider using this methodology for his or her empirical research? The reasons
not to adopt a discourse analytic approach are obvious. First, any new
Thus the task of discourse analysis is not to apply categories to participants'
method requires substantial investments of time and energy to master.
talk, but rather to identify the ways in which participants themselves actively
construct and employ categories in their talks. Further, all categoriz.ation Discourse analysis is certainly no exception to this rule, especially with the
is provisional; analysis requires constant reflexive attention to the process of relative shortage of methodological writings and established exemplars to
categorization of both the participant and the analyst. (Wood & Kroger, 2000, guide newcomers to the field. Second, and even more important, by defin-
pp.29-30) ition new methods are not institutionalized. Researchers therefore face
substantial barriers as they attempt to publish or present work that their col-
Even grounded theory although it seeks to generate categories from empir- leagues find unfamiliar and that can be diffrcult to relate to existing work in
ical findings, does not problematize them in the way that discourse analysis the field. Researchers who adopt the method face additional risks when
does. It accepts the researcher's "reading" of the data (subject to carrying out their work is evaluated for tenure or promotion because the relative rarity of
the necessary research protocols). Discourse analysts, on the other hand, are discourse analytic studies makes their evaluation difficult, and unfamiliar
attuned to the co-constfuction of the theoretical categories at multiple levels, reviewers may not appreciate their value. Third, discourse analysis is a
including researcher, research subject, academic community, and even labor-intensive and time-consuming method of analysi.s. Given the ever-
society, and they attempt to design and present their research in ways that ticking tenure clock and pervasive "publish or perish" cuhure in academia,
acknowledge these complex relationships (Alvesson & Sk6ldberg, 2000; there are easier and quicker alternatives for carrying out research.
Clegg & Hardy, 1996a; Hardy, Phillips, & Clegg, 2001). Despite these problems, we believe that there are many good reasons
The need to link text, context, and discourse, and to incorporate a highly why discourse analysis has an important role to play in tlte future of
subjective and reflexive use of research methods, poses a major chal- social science. These reasons outweigh the disadvantages of adopting a
lenge for researchers: How do we cope with all this complexity? We can new and relatively unproven research method and, at a personal level,
never study all aspects ofdiscourse, and we inevitably have to select a sub- they have convinced us to use discourse analysis in our own research and
set of texts for the purpose of manageability. Nonetheless, as discourse ana- led us to write this book to assist others who might want to use the
lysts we must still make reference to broader discourses, acknowledge the method.
 t2
13

In the remainder of this section, we outline five reasons why researchers

simply in what the social world means to the subjects who populate it; we
should consider using discourse analysis. Some of these reasons are specific
are interested in how and why the social world comes to have the meanings
to discourse analysis and theory itself, whereas others reflect the changing
that it does. We are also interested in how we, as researchers, are implicated
nature of our particular field of study----organization and management in that process (Clegg & Hardy, 1996a; Hardy et al., 2001). Discourse analy-
theory' The changing nature of the "organization" has resulted in a growing
sis, as one method for studying these more reflexive processes of social
need to find new ways of studying old topics, as well as effective
construction, is therefore attracting increasing attention (Alvesson &
approaches to studying new topics. At this point, it is worth adding a small
Kdrreman, 2000a).
disclaimer. Although we have written this book to be as general as possible,
As the linguistic turn has swept through disciplines, researchers have
and although we feel that the problems and solutions we discuss relate to a
turned to discourse analysis to study its implications for empirical research.
number of disciplines, many of our examples are organizational in nature.
Although somewhat late in adopting this view compared with the humanities
Yet insofar as organizational studies involves the study of individuals and
androther areas of social science, researchers in organization and manage-
societies, as well as organizations, the motivation behind the use of dis-
ment theory are also beginning to see language as increasingly important
course analysis in our field is not dissimilar to the reasons driving discourse
(Alvesson & Kiineman, 2000a).' The idea that organizations are socially
analysis in other fields.
constructed and exist primarily in language (broadly defined) is becoming
widely accepted. As a result, researchers are increasingly open to and inter-
The "Linguistic Tirrn"
ested in hnding new ways to examine these processes. Discourse analysis
over the last 30 years, a revolution of sorts has swept across the human- provides such a methodology because it is grounded in an explicitly con-
ities and social sciences. Beginning with the work of linguistic philosophers structionist epistemology that sees language as constitutive and construc-
such as wittgenstein (1967) and winch (1958), the idea thar language tive rather than reflective and representative (Wood & Kroger, 2000).
is much more thaf a simple reflection of reality-that, in fact, it is constitu-
tive of social reality-has become commonly accepted. This early work New and ReconceptualizedTopics of Study
heavily influenced sociologists such as Berger and Luckmann (1967) and
Broader changes in society have led to the emergence of new topics for
anthropologists such as Geertz (1973), whose work formed the foundation
study, which has reinforced the role of discourse analysis as a viable and
of a constructionist view of social phenomena. This view has permeated
useful research methodology. For example, the natural environment,
the social sciences and become well accepted as many disciplines are, in
globalization, and cultural studies have, relatively recently, captured the
Gergen's (1999, p. 16) terms, "pulsing toward the postmodern" and wrestling
interest of researchers in a number of disciplines who have made effective
with crises of representation and legitimation (Denzin & Lincoln, 1994; use of discourse analysis. Within the narrower confines of organization and
Rosenau, 1992).
management theory, the study of emotion (e.g., Fineman, 1996', Mumby &
The recognition of the constructive role of language problematizes the
Putnam, 1992) is one example of a relatively new area in which discursive
very nature of research as the objectivity, neutrality, and independence of
approaches have been applied to great effect. New topics of study raise new
tle researcher are called into question, as the nature of what passes for truth
challenges for researchers by creating new categories and drawing our
and knowledge is scrutinized, and as the question of how things work is
attention to how boundaries are constructed and held in place. Traditional
replaced by questions about what things mean (winch, 195g). The social
qualitative approaches may provide insight into the nature of these cate-
sciences are not only about counting--defining and measuring variables
gories, whereas quantitative research often allows generalizable claims to
and the relationships between them-they are also about interpreting what
be made about the relations between categories but neither helps us to
social relationships signify, to which a long history of qualitative research
understand how these categories came to be nor what holds them in place.
bears witness. with the linguistic turn, however, the demands of interpreta-
In fact, traditional methodologies often reify categories, making them seem
tive research are multiplied. As researchers, we are no longer interested
natural and enduring. Discourse analysis, on the other hand, provides a
 15
14

way of analyzing the dynamics of social construction that produce these Foucault, researchers have become interested in how processes of social
categories and hold the boundaries around them in place. construction lead to a social reality that is taken for granted and that advan-
Other subjects have been reconceptualized by researchers and now tages some participants at the expense of others (e'g., Clegg, 1989)' At the
require completely different approaches from those used in previous work. same time, researchers have sought to examine these political dynamics
For example, identity has long been a subject of study in a number of disci- without falling into the critical trap of "standing outside" the power rela-
plines, but primarily from an orientation in which researchers attempt to tions they are studying (Hardy & Clegg, 1996). These new challenges fac-
reveal or understand an individual's "true" or essential identity (see Nkomo & ing critical management studies have created a need for new methods that
Cox, 1996). More recently, discursive approaches are gaining ground in such expose the dynamics on which power distributions in organizations-and in
disciplines as psychology (e.C., Condor & Antaki, 1997; Potter & Wetherell, research-depend,
1987), gender studies (e.g., Tannen, 1994), organization and management This new and renewed concern with power has not only been confined
theory (e.g., Cal6s & Smircich, 1 991 ; Mumby & Stohl, I 99 I ; Wilson, 1996), to organization and management theory. Researchers in areas such as social
and social movement theory (e.g., Gamson, 1995) because of the insights *'ou"*"nt theory, communications, psycbology, and gender studies are
provided by an understanding of how identities are constructed on a continu- also increasingly attuned to the dynamics of power. As a result, there are
ous, interactive, discursive basis. significant opportunities for the application of such methods as critical
discourse analysis and critical linguistic analysis (Fairclough, 1992' 1995
The Revitalization of Critical Management Studies Mumby, 2000; Mumby & Stohl, 1991; Parker, 1992) to a variety of settings,
in addition to those related to organizations.
An important reason for the growing appeal of discourse analysis in
organization and management theory derives from the renewed interest in
The Development of Postbureaucratic Organizational Forms
critical management studies. Critiques of managerialism have a long-
standing tradition in organization and management theory as a result of Another reason specific to the increased use of discourse analysis in
early Marxist traditions and more radical readings of Weber (Hardy & organization and management theory is the changing nature of organiza-
Clegg, 1996). They appear in a variety of theoretical streams such as labor tional and management practice over the last few decades. In reflecting on
process theory (e.g., Braverman,l9T4: Burawoy, 1979; Edwards, 1979), the last 30 years of organization and management theory, Clegg and Hardy
work on power (e.g., Benton 1981; Clegg, 1975; Hardy, 1985; Lukes, . (1996b, p. 2) noted that in the 1960s, "hierarchies were the norm, personal
1974), and studies of culture and ideology (e.g., Smircich, 1983; Weiss computers had not been invented, and the only mode of instantaneous com-
and Miller, 1987; Willmott, 1993), to name but a few. The advent of post- munication was the telephone. ifhe new technologies that were to challenge
modernism in organization and management theory initially challenged radically accepted organization designs seemed unthinkable'" Today, we
this line of thinking (e.g., Bunell, 1988; Cooper & Burrell, 1988). Over witness an iuray of new organizational forms; the widespread acceptance of
time, however, the integration of postmodern and poststructuralist new information technologies; the increasing globalization of business,
insights has reinvigorated critical management studies and attracted a trade, and culture, as well as resistance to it; and the increasing importance
number of researchers to what is a revitalized agenda in critical manage- of knowledge- and symbol-intensive firms.
ment studies (Alvesson & Deetz,2000; Alvesson & Willmott, 1992a, ,,^,-
o "'dpitliitiirul
to slyqy 1l19_ngry
1992b: Fournier & Grey, 2000),'? aspects of organizations. It is increasingly difficult to study organ-
Much of the renewed research attention has focused on the intersection " "-ate "f
between critical and postmodern theory (Alvesson &Deetz, 1996; Mumby, theii fluid and contradictory dynamics. As a result lve search for thegories,
1992) and, specifically, on the connection between power and meaning- narratives, and ryn'rlgls-the at_scogrsg!-that h contra-
the way in which knowledge is bound up in the dynamics of power dictory flows and make them "real" for us (Chia, 2000). Discourse analysis
(e.g., Knights & Morgan, 1991). Building particularly on the work of provides a powerful way to study these slippery, ephemeral phenomena
16
t7

and, as such, is vital if we are to inform and be informed by organizational a range of literatures and show some of the different ways in which
and management practice. discourse analysis can be used. Third, we have used the writing of this
book to reflect in some depth on our own work and identify a number of
The Limits of Traditional Methods and Theories
challenges that researchers face as they adopt this method. Finally, we
The final reason we believe that discourse analysis is important grows hope to offer some suggestions, based on our own experiences, of how to
out of the increasing calls for pluralism that can be heard across the social tackle these challenges.
sciences (e.g., Kaghan & Phillips, 1998). The idea of ,,one best method" has The remainder of this book is organized in the following way. In
been challenged more and more frequently; in fact, it has largely been Chapter 2, we provide a framework for understanding the different forms
replaced by the idea that research is best served by a plurality of methods of discourse analysis that currently appear in the literature in a variety of
and theories (Clegg & Hardy, 1996b). Many researchers have begun to find disciplines. We begin by focusing on the theoretical assumptions underly-
traditional approaches to research too limiting and repetitive. Rather than ing approaches to discourse analysis and then discuss some of
$ifferent
using the same method to study the same phenomenon more intensively, the the range of empirical topics that have been explored using discourse
use of a very different method can provide far more insight (Alvesson & analysis. Our intention in Chapter 2 is to provide the reader with a frame
Deetz, 2000). Using a nontraditional method provides a way to see things for understanding approaches to discourse and an appreciation of the
that have been obscured by the repeated application of traditional methods- range of potential topics that can be studied using discourse analysis. In
all ways of seeing are also ways of not seeing. using a discursive approach Chapter 3, we introduce the reader to our own work in this area. We out-
can allow researchers to build on and complement other bodies of theor- line a number of studies we have conducted and explain the types of
etical work by introducing new ideas, new concepts, and new challenges. choices we made in terms of data, data analysis, and general theoretical
There is also the fact that it can be more interesting to use less traditional orientation. We also discuss the contributions that discourse analysis
methods to study the world of organizations. They are, by definition, less made to our understanding of the phenomena that we studied. In Chapter 4,
institutionalized, which allows researchers to use more creativity in their we explore the question of how to do discourse analysis. Drawing on our
application and more innovation in their interpretation. experience in carrying out the research program outlined in preceding
In summary, we see discourse analysis as an important contribution to chapters, we identify some of the key challenges facing researchers
increasing plurality in research, a way to incorporate the linguistic turn and embarking on a research project and discuss some of the ways in which
to study new phenomena and practices, as well as to reinvigorate agendas of they might be addressed. In our final chapteq we sum up our current
critical theory. It may pose problems, and old certainties may well dis- thoughts on discourse analysis as a f,reld and point to some of the major
appear but, as Clegg and Hardy (1996b, p. 8) pointed out, ,,It is in the sruggle hurdles that we still need to clear.
between different approaches that we leam, and from the diversity and ambi-
guity of meaning; not through the recitation of a presumed uniformity, con- Notes
sensus, and unity, given in a way that requires unquestioning acceptance."
1. The degree to which social constuctivism is accepted in organization and
management theory varies geographically. In Europe, few researchers would have
What Lies Ahead difficulty with the basic premises of social constructivism. In North America, it is
less accepted, although this is changing rapidly.
This book sets out to help aspiring discourse analysts in four ways. First, 2. For example, critical management studies has been established as an integral
we provide a coherent framework for understanding the different forms part ofthe preconference professional development workshops at the Academy of
of discourse analysis that currently appear in the literature, Second, we Management in the United States, and plans exist to apply for interest group status.
In the United Kingdom, the flust critical management studies conference was held
present a wide range of empirical studies that have been conducted across
in 1999.