22

CHAPTER

LANGUAGE IN QUALITATIVE
RESEARCH
Introduction 483 Rhetorical analysis 491

Discourse analysis 483 Conversation analysis 493

Main features of discourse analysis 484 Overview 496
Interpretive repertoires and detailed procedures 486
Critical discourse analysis 488 Key points 497

Narrative analysis 489 Questions for review 497

CHAPTER OUTLINE

This chapter is concerned with approaches to the analysis of language, including discourse analysis. The
chapter explores

• discourse analysis, which examines how language is used and the effects on understanding reality that
are accomplished through this—such approaches include critical discourse analysis, where there is an
emphasis on understanding the effects of historical and social context on the type of language that is
produced;

• two further approaches that focus on specific forms of language and their effects in creating
meaning—narrative analysis (also sometimes referred to as storytelling research), and rhetorical
analysis, which concentrates on the use of specific forms of language to convince or persuade;

• and finally, conversation analysis, which is a fine-grained approach to analysing the detail of how
language is used, especially in naturally occurring conversations.
 Discourse analysis 483

Introduction
Language is bound to be of importance for organizational significantly more than a medium through which business
researchers. It is, after all, through language that we ask research is conducted (such as by asking questions in inter-
people questions in interviews and through which the views). Instead language is treated as a focus of attention in
questions are answered. Language is also central to the its own right. This type of research is broadly c­ onstructionist
structuring of organizations, if only because people in in orientation, hence language is not just seen as reflective
work organizations rely so heavily on talk—in meetings, of what goes on in an organization; instead, it is a way of
on the telephone, in the cafeteria—to accomplish their constructing particular understandings of phenomena.
everyday business. It is through language that people This means, for example, that as soon as managers in a
in organizations exchange information, skills, services, public-sector organization start to talk of their clients as
and resources and make sense of their situation through ‘customers’, a whole new way of defining the organization’s
interaction with each other. In managerial work a great purpose and activities is introduced.
deal of emphasis is placed on verbal interaction or talk, The first part of the chapter is concerned with what
as numerous research studies highlight. For example, may be characterized as meso-level (Alvesson and Kär-
Mintzberg’s (1973) study reports that verbal contacts, reman 2000) or ‘context-sensitive approaches’ (Grant et
face-to-face and on the telephone, accounted for 75 per al. 2004) to the study of language which take account of
cent of senior managers’ time and 67 per cent of their factors beyond the text itself. In addition to being sensi-
activities (see Chapter 13 for an explanation of Mint- tive to the context in which language is produced, these
zberg’s study). Other studies have shown that between approaches are concerned with finding generalizable
57 and 89 per cent of managerial time is spent in verbal patterns and going beyond the detail of the text. They
interactions of one kind or another (Boden 1994). The therefore explore social and historical context, and other
role of the business researcher who focuses on language factors that influence how language is produced, dis-
is to explore the nature of the relationship between lan- seminated, and consumed. Discourse, including critical
guage and action in these instances. discourse analysis, and to a lesser extent narrative and
What is crucial about the approaches discussed in this rhetorical analysis, can be considered context-sensitive
chapter is that, unlike other methods that use language approaches. The final part of the chapter deals with a
in business research, they treat language as a topic rather more fine-grained approach to analyzing language use,
than as a resource. This means that language is treated as in the form of conversation analysis.

Discourse analysis
While discourse analysis is not the only way of analyzing The first part of this chapter will outline the main fea-
language qualitatively, it represents a widely used method tures associated with a particular type of discourse analy-
in business research. There is an international conference sis which has evolved its own technical vocabulary and
dedicated to discursive approaches to organizational set of techniques. This approach is associated with such
analysis as well as a handbook (Grant et al. 2004) and a writers as Potter (1997), Potter and Wetherell (1987,
journal, Discourse and Society. As researchers who study 1994), Billig (1992), and Gilbert and Mulkay (1984); it
discourse note, the focus on language in business research is suggested to be characterized by two distinctive fea-
is not surprising because language plays such a significant tures at the level of epistemology and ontology (Potter
role in ‘constructing, situating, facilitating and commu- 1997) (see also Key concept 22.4).
nicating the diverse cultural, institutional, political and
socio-economic parameters of “organizational being”’ • It is anti-realist: in other words, it denies that there is
an external reality awaiting a definitive portrayal by
(Grant, et al. 1998: 12). A key strength of discourse analy-
the researcher and it therefore rejects the notion that
sis concerns its ability to be applied to naturally occurring
any researcher can arrive at a privileged account
conversation and documents as well as other forms of
of the aspect of the social world being investigated.
written and spoken language.
484 22 Language in qualitative research

22.1 KEY CONCEPT

What is discourse analysis?
According to Potter (1997), discourse analysis ‘emphasizes the way versions of the world, of society, events and
inner psychological worlds are produced in discourse’ (1997: 146). This definition of discourse analysis means
that discourse is not just a mirror on the social world around us but in many ways plays a key role in producing that
world. How we say things—our phrases, our emphases, the things we leave out—is meant to accomplish certain
effects in others. In so doing, we have an impact on others’ perceptions and understandings and as such on their
and our reality.

Some discourse analysts, however, adopt a stance that 3. What resources are available to perform this activity?
is closer to a realist position, but most seem to be anti-
(Potter 2004: 609)
realist in orientation.
The consideration of discourse analysis that follows
• It is constructionist: in other words, the emphasis is
placed on the versions of reality propounded by mem- draws on two studies:
bers of the social setting being investigated and on the • research into the discourses applied to unemployed
fashioning of that reality through their renditions of it older workers in Australia (Ainsworth and Hardy
(see Key concepts 2.7 and 22.1). More specifically, the 2009);
constructionist emphasis requires a recognition that
discourse entails a selection from many viable rendi-
• a study of MBA students’ use of role models in profes-
sional identity formation (Kelan and Mah 2014).
tions and that in the process a particular depiction of
reality is built up. The first study, by Ainsworth and Hardy (2009; see
Research in focus 22.2), shows how discourses of the
Thus, discourse is not simply a neutral device for impart-
mind and body are used to discipline unemployed older
ing meaning. People seek to accomplish things when
workers in a way that is disempowering; the second study,
they talk or when they write; discourse analysis is con-
by Kelan and Mah (2014; see Research in focus 22.3),
cerned with the strategies they employ in trying to cre-
provides an illustration of how the notion of i­ nterpretative
ate different kinds of effect. This version of discourse
repertoire can be used to understand gendered processes
analysis is, therefore, action orientated—that is, it is a
of identity formation among MBA students.
way of getting things done. This is revealed in three basic
discourse-analytic questions:
Main features of discourse analysis
1. What is this discourse doing?
2. How is this discourse constructed to make this 1. Not just speech. Discourse analysis is an approach that
happen? can be applied to forms of communication and not

22.2 RESEARCH IN FOCUS

The application of mind and body discourses
to older workers
Ainsworth and Hardy (2009) studied the discourses pertaining to the employment of older workers through a
discursive analysis of a parliamentary inquiry in Australia, which was set up to examine the barriers that older
unemployed workers face in regaining employment. The inquiry lasted just over a year (1998–9) and comprised a
series of written submissions and public hearings held throughout Australia, where employee and lobby groups,
unions, government representatives, and private individuals were invited to appear. A final report was then
 Discourse analysis 485

published in 2000 on the findings of the inquiry. One of the reasons for selecting the inquiry as a research site was
because of the accessibility, volume, and range of available texts relating to it, which included media releases,
written submissions, and more than 1000 pages of oral evidence and testimony relating to the public hearings,
which had been transcribed verbatim. There was the further advantage that many of these documents could be
obtained from the government website. Ainsworth and Hardy (2009) claim that, because these ‘naturally
occurring’ texts were generated independently of the researcher, they have the advantage of not being subject to
reactive effect and provided very useful materials for systematic analysis.

The authors identify two discourses in the texts:

1. physical discourses relating to the body, which portray ageing as a process of inevitable decline;
2. discourses of the mind, which psychologize and individualize the problem of unemployment.

They argue that, while these discourses have separate effects, their consequences for identity formation are even
greater when brought together through a normative ‘mechanism of grief’, which encourages older unemployed
workers to make use of labour market interventions that help them to accept their loss of employment rather than
locate permanent job opportunities. Unemployed older workers were thus advised to ‘manage the grief’ associated
with their loss of employment and to ‘resolve the anger’ that they felt in relation to job loss because this was
deemed ‘unhealthy’ and claimed to be harming their chances of re-employment. The personal stories told by
unemployed older workers were thereby discursively regulated through retelling and reframing by others in a way
that deflected responsibility away from government agencies or employers. The authors conclude that, ‘rather than
provide space for resistance, the intersection of these discourses disempowers an already disadvantaged group’
(Ainsworth and Hardy 2009: 1200).

22.3 RESEARCH IN FOCUS

Interpretative repertoires in the identification
of role models by MBA students
Kelan and Mah (2014) were interested in how MBA students construct a professional self-identity through
identification of gendered role models. Their interest in this topic arose from previous research findings which
suggested that women managers struggle to find examples of women leaders with whom they can identify. They
conducted 20 in-depth interviews with full-time MBA students at an elite British business school, ten of whom
were men and ten women, each interview lasting between 45 and 110 minutes. The researchers also used visual
methods to encourage interviewees’ engagement with this topic; their questions related to a person the interviewee
admired, as follows:

We asked you to bring along a picture of a person in business you admire. Who have you chosen? What do
you admire in this person?
(Kelan and Mah, 2014: 94)

Although only half of the interviewees brought a photo of a person they admired, by asking interviewees to bring
along a picture to the interview, the researchers were able to encourage interviewees to think about this in advance.
All the male interviewees selected men as the people they admired, whereas the female interviewees selected men
and women. Through their analysis, Kelan and Mah identified two interpretative repertoires, comprised of common
tropes which interviewees used to talk about the person they admired. They refer to these as the ‘idealization’
repertoire and the ‘admiration’ repertoire. They found significant differences between men and women MBA
students in their study in terms of the way students used these two interpretative repertoires to construct a
professional self-identity. The idealization repertoire was used by all the male students, whereas the repertoire of
admiration was used almost exclusively by women students when talking about individuals they admired.
486 22 Language in qualitative research

only to spoken language or talk. This involves treating conversation or article in one way rather than in an-
texts (see Chapter 23) as interrelated to each other other way, is a crucial component of seeing discourse
and dependent on context. The types of texts that as a solution to a problem. For example, Ainsworth
may be analysed using a discourse analytic approach and Hardy (2009; see Research in focus 22.2) argue
are wide ranging, and include corporate annual re- that discourse shapes the rules that determine how
ports, government inquiries into organizational di- we speak and act in relation to a given topic, in a way
sasters, and the content on organizational websites. that gives certain actors more legitimacy and rights to
Moreover, in discourse analysis there is much less of commentate than others. They found that discourses
an emphasis on naturally occurring talk, so that talk that represent the physical process of human ageing
in research interviews can be a legitimate target for as a process of inevitable decline and discourses of
analysis. the mind that focus on psychologizing feelings of
2. Contextual understanding. Discourse analysts have a loss and anger associated with job losses were used
preference towards locating the situational specifics to marginalize older workers and to exclude them
of talk in the context of their occurrence. As Potter from the labour market. Formulating understand-
(1997: 158) puts it, discourse analysts prefer to avoid ings of unemployment in this way thus discourages
making reference in their analyses to what he refers to collective acknowledgement of responsibility for the
as ‘ethnographic particulars’ and argues that instead problem of older worker unemployment and invites
they prefer ‘to see things as things that are worked older unemployed people to accept their disempow-
up, attended to and made relevant in interaction ered situation. This example illustrates the potential
rather than being external determinants’. However, for discourses to affect power relations between social
discourse analysis practitioners are less committed actors, a point to which we will return to in the section
to this principle than conversation analysts, in that on critical discourse analysis that follows.
the former sometimes show a greater preparedness
to make reference to ‘ethnographic particulars’.
3. Resists codification. Discourse analysts are opposed to
Interpretive repertoires and detailed
the idea that their analytical practices can be codi- procedures
fied and argue that such a codification is probably Potter and Wetherell (1994) highlight two tendencies
impossible. One useful point of departure for dis- within discourse analysis, although they acknowledge
course analysis research that has been suggested by that the distinction is somewhat artificial. One is the
Gill (1996), following Widdicombe (1993), is to treat identification of ‘the general resources that are used to
the way that something is said as being ‘a solution construct discourse and enable the performance of par-
to a problem’ (Widdicombe 1993: 97, quoted in Gill ticular actions’ (1994: 48–9), and is concerned with iden-
1996: 146). Gill (2000) also suggests adopting a pos- tifying interpretative repertoires. The other is concerned
ture of ‘sceptical reading’. This means searching for to identify ‘the detailed procedures through which ver-
a purpose lurking behind the ways that something is sions are constructed and made to look factual’ (1994:
said or presented. 49). We will now explore these two strands of discourse
4. Sensitivity to what is unsaid. A further feature to be analysis.
aware of is that what is said is always a way of not To illustrate the idea of an interpretative repertoire, we
saying something else. In other words, either total will refer to the study of professional role models adopted
silence on a topic, or formulating an argument in a by MBA students described in Research in focus 22.3.

TIPS AND SKILLS

Using existing material
As some of these examples of discourse analysis illustrate, you may be able to use the technique to study an
issue of interest to you based on materials that are publicly available, such as speeches. This means that you do
not have to collect new data and you can devote more attention analysing the materials. Research in focus 22.2
provides an example of a study that relied exclusively on publicly available written texts.
 Discourse analysis 487
Kelan and Mah (2014: 93) define the interpretative reper- really exist. The notion of the interpretative repertoire
toire as ‘a repeatedly encountered construction employed is interesting because it brings out the idea that belief
in sensemaking. It is a repertoire because only a limited and action take place within templates that guide and
number of terms are used’. They go on to say that these influence the writer or speaker. However, the interpreta-
common-sense units of understanding have ‘an “off-the- tive repertoires identified by Kelan and Mah (2014) by
shelf” character and can be used flexibly to make a point no means exhaust the range of possibilities for analysis,
in any given situation’. They identify two interpretative as the advantages of the notion of interpretative reper-
repertoires that reflect the ways that individuals talked toires stem primarily from its flexibility in accounting
about people in business whom they admire. for a diverse range of social practices. Hence, Potter and
Wetherell (1987) suggest that repertoires are available
• The idealization repertoire. This is comprised of
to people with many different social group memberships.
respondents’ talk about people who are business
They also point out that there is no need to attempt to
founders or very senior in organizations, such as CEOs.
find consensus with regard to repertoires—because they
Themes include ‘being self-made’, having achieved
are used to perform different sorts of accounting tasks,
success through hard work and perseverance, and
individuals are able to draw upon a variety of repertoires
‘being authentic’. This is illustrated by one of their
in different situations. Finally, they emphasize that ‘the
respondents, Luke, who talks about the person he
concept of repertoire is but one component in a system-
admires in these terms, stating:
atic approach to the study of discourse’ (Potter and Weth-
He got a football scholarship to school (…) and started erell 1987: 157), one that in a few years’ time may be
up a load of car dealerships. And made kind of five or six developed further or even discarded.
hundred million dollars by the time he was fifty (…) it’s In discourse analysis there is also an emphasis on the
that sort of completely self-made, incredibly sharp (…) resources that are employed in conveying allegedly fac-
he’s got very high levels of integrity, you know, what you tual knowledge—or what Potter and Wetherell (1994)
see is what you get. might describe as quantification rhetoric, by which is
(Kelan and Mah 2014: 96) meant the ways in which numerical and non-numerical
statements are made to support or refute arguments.
While the themes of being self-made and being authentic
Instead, the texts largely consist of general statements,
might seem contradictory, Kelan and Mah suggest that
claims, and conclusions. This is interesting, given the
they justify and balance the two ideals of ‘wealth, sta-
importance of quantification in everyday life and the
tus and power’ and ‘integrity, passion and being true to
tendency for many social scientists to make use of this
the self’. They then go on to contrast this with a second
strategy themselves (John 1992).
repertoire.
A number of further characteristics apply to discourse
• The admiration repertoire. This is characterized by analysis, the most important of which are presented below.
ambivalence and caveats in talking about the admired
person, such as the ‘Superwoman’ who appears to do • Reading the detail: discourse analysts incorporate the
it all. This includes talk about their not-admired char- conversation analysis preference for attention to the
acteristics (neutral or negative), as well as the positive details of discourse.
traits of the individual. It thereby involves more criti- • Looking for rhetorical detail: attention to rhetorical
cal evaluation of the person with whom one identifies. detail entails a sensitivity to the ways in which argu-
For example: ments are constructed.
Frances: ‘I like her’ cause she’s a woman who’s basically • Looking for accountability: discourse analysts draw on
lived her life according to her own values (…) I was always conversation analysis practitioners’ interest in and
a bit suspicious that she’d be a bit of a fraud. Until I actually approach to accounts. From the point of view of both
saw her speak (…) And then I realized that she’s just very, conversation analysis and discourse analysis, dis-
very open, you know, and what you see is what there is. course can and should be regarded as accounts. For
(Kelan and Mah 2014: 98) discourse analysis practitioners, the search for
accountability entails attending to the details through
A key difference, then, between the second and the first
which these accounts are constructed.
interpretative repertoire is that the admiration repertoire
contains much greater ambivalence, reflecting tensions • Cross-referencing discourse studies: Potter and Weth-
erell suggest that reading other discourse studies is
and suspicions about whether a person such as this could
488 22 Language in qualitative research

itself an important activity. First, it helps to sharpen In an organizational context, one of the things that criti-
the analytic mentality at the heart of discourse analy- cal discourse analysis practitioners seek to trace is how
sis. Secondly, other studies often provide insights that discourses are constructed and maintained in relation
are suggestive for one’s own data. to certain phenomena, such as globalization or strategic
management. Analysis seeks to reveal the meaning of a
This approach to discourse analysis has been criticized
phenomenon by exploring how
for being too narrow in focus or not sufficiently sensitive
to context. The anti-realist inclination of some discourse
• the discourse has come to have a particular meaning
analysis practitioners has been a source of controversy. It today, when 40 or 50 years ago it might have had none
has been claimed that the emphasis on representational or a quite different meaning;
practices through discourses sidelines any notion of a
pre-existing material reality that can constrain individual
• the discourse draws on and influences other dis-
courses;
agency (Thompson and Harley 2012). Reality becomes lit-
tle more than that which is constituted in and through dis- • the discourse is constructed through texts (such as
course. This lack of attention to a material reality that lies academic articles or journalistic writing);
behind and underpins discourse has proved too abstracted • the discourse gives meaning to social life and makes
for some social researchers and theorists. This is an issue certain activities possible, desirable, or inevitable;
that we will deal with next, when we examine critical
• particular actors draw on the discourse to legitimate
discourse analysis. The main point to note at this stage is their positions and actions (Phillips and Hardy
that, while many discourse analysis practitioners are anti- 2002: 8).
realist, an alternative realist or critical realist position in
relation to discourse is also feasible (Fairclough 2005). As the second point in the above list indicates, discourses
are conceived of as drawing on and influencing other dis-
courses. So, for example, the discourse of globalization
Critical discourse analysis might affect discourses on new technology, free trade and
Critical discourse analysis is principally concerned with liberalism, or corporate social responsibility. However, this
capturing and analysing how language is used in specific is not always a complementary process, as in some cases
socio-historical contexts in order to generate particular discourses compete with each other for dominance in what
effects. It also emphasizes the role of language as a power is termed dialogical struggle (Keenoy et al. 1997). An exam-
resource. This approach is associated with writers such ple of this can be seen in the analysis by Legge (1995) that
as Fairclough (1992, 1995, 2003), Hardy (2001), and traces the changing rhetorics of personnel management
Phillips and Hardy (2002). Critical discourse analysis and HRM in the UK. Legge argues that ‘the importance of
additionally draws on the writings of the social theorist HRM, and its apparent overshadowing of personnel man-
Michel Foucault (1974, 1979, 1980), whose work uncov- agement, lies just as much and (possibly more so) in its
ers the representational properties of discourse as a vehi- function as rhetoric about how employees should be man-
cle for the exercise of power. Foucault draws attention aged to achieve competitive advantage than as a coherent
to the disciplinary practices that enable particular ver- new practice’ (1995: xvi). This has the potential to give
sions of subjectivity to be constructed in different socio- rise to a rhetoric–reality gap, in which discourses coexist
cultural moments, and the role of language within this. and are translated into social practice in a variety of ways
Consequently, the notion of discourse is broader than in (Watson 1994a). Critical discourse analysis thus involves
other forms of discourse analysis, as this summary by exploring why some meanings become privileged or taken
Phillips and Hardy (2002) highlights. for granted and others become marginalized. In other
words, discourse does not just provide an account of what
We define a discourse as an interrelated set of texts, and
goes on in organizations; it is also a process whereby mean-
the practices of their production, dissemination, and re-
ing is created. This involves asking ‘who uses language,
ception, that brings an object into being (I. Parker 1992)
how, why and when’ (Van Dijk 1997: 2).
… In other words, social reality is produced and made real
Analysis of a discursive event is usually carried out
through discourses, and social interactions cannot be fully
according to a ‘three-dimensional’ framework, which
understood without reference to the discourses that give
them meaning. As discourse analysts, then, our task is to
proceeds as follows:
explore the relationship between discourse and reality.
• examination of the actual content, structure, and
(Phillips and Hardy 2002: 3) meaning of the text under scrutiny (the text dimension);
 Narrative analysis 489
change in discourse’ (2005: 931). Fairclough identifies
• examination of the form of discursive interaction used
to communicate meaning and beliefs (the discursive four sets of organizational research issues that a critical
practice dimension); realist approach to discourse analysis can address:

• consideration of the social context in which the discur- • emergence: founded on the notion that ‘new’ organiza-
sive event is taking place (the social practice dimension) tional discourses emerge ‘through “reweaving” rela-
(Grant et al. 2004: 11). tions between existing discourses’ (2005: 932);

A further key concept within critical discourse analysis is • hegemony: focusing on how particular discourses
the notion of intertextuality, which draws attention to the become hegemonic in particular organizations and on
notion of discourse as existing beyond the level of any par- ‘how discourse figures within the strategies pursued
ticular discursive event on which analysis is focused. The by groups of social agents to change organizations in
notion of intertextuality thus enables a focus on the social particular directions’ (2005: 933);
and historical context in which discourse is embedded. • recontextualization: involving identification of the
As noted earlier, there has been some criticism of dis- principles through which ‘external’ discourses are
course analysis for apparently ignoring material reality internalized within particular organizations;
which exists separately from the discursive realm (Reed
• operationalization: focusing on how discourses are
2000; Thompson and Harley 2012). A strong critic of dis- operationalized, transformed into new ways of acting
course analysis on this basis has been Fairclough, who has and interacting, inculcated into new ways of being, or
argued that it should be developed in a direction which materialized, within organizations.
adopts the tenets of critical realism. Fairclough is scepti-
cal of the anti-realist assumptions of some discourse ana- Discourse analysis has also been criticized because what
lysts who reject objectivist conceptions of organization different researchers understand the term ‘discourse’ to
as social structure in favour of seeing it as ‘an interactive mean varies considerably, and so does their approach to
accomplishment’ (2005: 917), according to a construc- analysis. There is thus a danger, noted by Alvesson and
tionist perspective (see Chapter 2). He quotes Mumby Kärreman (2000), that the term ‘discourse analysis’ is
and Clair (1997) as typical of the latter position in saying too broad to be meaningful, authors treating the term
‘we suggest that organizations exist only in so far as their as though it has a clear, broadly agreed-upon meaning,
members create them through discourse’ (1997: 181). which, just from reading this chapter, you will be able to
Instead, Fairclough recommends an approach that see it does not. Consequently, ‘discourse sometimes comes
centres on the tension between organizational discourse close to standing for everything, and thus nothing’ (Alves-
and organizational structure. Therefore, a critical realist son and Kärreman 2000: 1128). Building on their earlier
approach to discourse analysis involves analysing not just article, Alvesson and Kärreman (2011) also warn that the
the discourse per se but also its relationship to non-dis- privileging of discourse, and primarily of talk and language
cursive elements. This is particularly important in rela- as the central force in organizational meaning-making,
tion to the study of organizational change because, ‘while contributes towards a marginalization of the non-discur-
change in discourse is a part of organizational change, sive, which includes the material, embodied, and unarticu-
and organizational change can often be understood lated. This criticism does not mean that discourse analysis
partly in terms of the constructive effects of discourse should not be employed, of course, but rather that when
on organizations, organizational change is not simply working in this tradition it is important to recognize some
of the pitfalls and omissions which critics have identified.

Narrative analysis
Narrative analysis is an approach to the elicitation and to the collection and analysis of data neglect the fact that
analysis of language that is sensitive to the sense of tem- people perceive their lives in terms of continuity and
poral sequence that people, as tellers of stories about process; attempts to understand social life that are not
their lives or events around them, detect in their lives and attuned to this feature neglect the perspective of those
surrounding episodes and inject into their accounts. Pro- being studied. Life history research (see Chapter 20) is
ponents of narrative analysis argue that most approaches an obvious location for the application of a narrative
490 22 Language in qualitative research

analysis, but its use can be much broader than this. of organizations as a collective storytelling system. He
Mishler (1986: 77), for example, has argued for greater gives an example of a strategic planning session where
interest in ‘elicited personal narratives’. In his view, and the CEO of the company uses stories to explain how the
that of many others, the answers that people provide in printing industry has changed and to convey a sense of
qualitative interviews can be viewed as stories that are things being better now than they used to be. Boje argues
potential fodder for a narrative analysis. In other words, that through this the CEO gains political advantage by
narrative analysis relates not just to the life span but also portraying the current context as more favourable than
to accounts relating to events and the interconnections the past one. However, not all organizational stories
between them. Some researchers apply narrative analy- are as coherent as this example implies. The concept of
sis to interview accounts (e.g. Riessman 1993), while ‘microstoria’, or ‘little stories’, is used to refer to the telling
others deliberately ask people to recount stories (e.g. R. of fragmented, terse stories. Boje et al. (2016) conducted
L. Miller 2000). Coffey and Atkinson (1996) argue that an historical archive analysis (see Chapter 23) into the
a narrative should be viewed in terms of the functions Burger King Corporation (1978–2015) to explore the
that the narrative serves for the teller. The aim of nar- role of alternative and competing ‘antenarratives’ in
rative interviews is to elicit interviewees’ reconstructed stakeholders’ retrospective and prospective sensemaking
accounts of connections among events and between about strategic change. Basing their study on analysis of
events and contexts. documents such as corporate annual reports, the authors
In business research, the study of narratives in orga- argue that this narrative method draws attention to sto-
nizations is well established (see Key concept 22.4 for a ries that sometimes ‘speak against organizational power
definition). Such analyses generally entail exploration of in situated context’ (2016: 401). This more complex
‘spoken or written accounts of connected events’ (Boje et approach acknowledges that while narrative analysis
al. 2016: 392), which are used to generate insight into often focuses on ‘relatively coherent narratives that have
organizational sensemaking processes (Weick 1995). In a clear beginning and an end’ (Vaara et al. 2016: 496),
a now classic study, Boje (1991) analyses the types and this is not a necessary requirement for narrative analysis
uses of stories in an office supply firm to develop a theory (see Key concept 22.4).

22.4 KEY CONCEPT

What are organizational narratives?
Vaara et al. (2016) suggest that organizational narratives have at least six key features.

1) Organizational narratives are ‘temporal, discursive constructions that provide a means for individual, social and
organizational sensemaking and sensegiving’ (2016: 498). They are associated with other language-based
approaches to analysis, including discourse and rhetorical analysis. However, Vaara et al. argue that it is the
temporal aspect of narratives which makes them distinct from these other approaches. This makes them
particularly suited to use in the study of change.
2) Organizational narratives are ‘not often fully fledged stories or accounts’ (2016: 498), in the way that classical
literary theory suggests. Instead their plotlines remain implicit and they are often fragmented rather than fully
formed or consensually agreed upon.
3) Analysis of organizational narratives requires a focus on how they are produced and consumed. The authors
use the term ‘story’ to refer to ‘existing narratives that can be told and retold in various forms’ (2016: 498).
4) Organizational narratives are part of multifaceted structures. At the macro level they are linked to societal
narratives (in a similar way to discourse analysis), while at the micro level they are comprised of discursive and
rhetorical elements which ensure their spread or appeal.
5) While organizational narratives are usually associated with written or spoken language, they also relate to other
modes of communication such as visual or audio.
6) Organizational narratives fulfil a key function in processes of organizational stability and change. Specifically, they
argue that narratives have both performative power, through providing accounts of events that frame them as
either change or stability, and agency, the potential to influence how events unfold or reproducing the status quo.
 Rhetorical analysis 491

22.5 RESEARCH IN FOCUS

An example of narratives in a hospital
Brown (1998) has examined the competing narratives involved in the aftermath of the introduction of a hospital
information support system (HISS) at a British hospital trust referred to as ‘the City’. The IT implementation was
largely seen as unsuccessful because of cost overruns and the absence of clear clinical benefits. Drawing on his
interviews with key actors regarding the IT implementation and its aftermath, Brown presents three contrasting
narratives—the ward narrative; the laboratory narrative; and the implementation team’s narrative—thereby
presenting the perspectives of the main groups of participants in the implementation.

The three contrasting narratives provide a very clear sense of the organization as a political arena in which groups
and individuals contest the legitimacy of others’ interpretations of events. Thus, ‘the representations of each
group’s narrative are described as vehicles for establishing its altruistic motives for embarking on the project, and
for attributing responsibility for what had come to be defined as a failing project to others’ (Brown 1998: 49).

Thus, while the three groups had similar motivations for participating in the initiative, largely in terms of the
espousal of an ethic of patient care, they had rather different latent motivations and interpretations of what went
wrong. In terms of motivations, whereas the ward narrative implied a latent motivation to save doctors’ and nurses’
time, the laboratory team emphasized the importance of retaining the existing IT systems, and the implementation
team placed the accent on the possible advantages for their own careers, in large part by the increased level of
dependence on their skills. In terms of the contrasting narratives of what went wrong, the ward narrative was to do
with the failure of the implementation team to coordinate the initiative and meet deadlines, and the laboratory
team emphasized the tendency for the implementation team not to listen or communicate. As for the
implementation team, their diagnosis was to do with the ward staff failing to communicate their needs, lack of
cooperation from the laboratory staff, and poorly written software.

Narrative analysis is suggested to be particularly suited was that it enabled exploration of the organization as
to the study of organizational culture and change. This an arena where a variety of perspectives and viewpoints
approach is also useful in studying issues of power and coexist, rather than a monolithic entity with a single
politics in organizations, as indicated by the study in voice. This is an example of a ‘composite narrative’ that
Research in focus 22.5. As Brown (1998) notes, one of captures the collective meanings held by a group of orga-
the advantages of using narrative analysis in this study nizational members (Vaara et al., 2016).

Rhetorical analysis
Related to narrative analysis is an approach that audiences. For example, rhetorical analysis has been
focuses on the importance of rhetorical devices as a used to critique management fashions and manage-
means of communication and persuasion within man- ment gurus by exploring how language is used to com-
agement and organization. This includes analysis of municate ideas to global audiences (B. Jackson 2001).
classic rhetorical devices, such as argumentation, as It is also applied in the study of leadership, as another
well as various literary devices, including tropes such organizational context in which language is targeted
as metaphor, synecdoche, metonymy, and irony. Rheto- at large audiences. For example, the study by Herac-
ric and tropes are argued to be an unavoidable feature leous and Klaering (2014) described in Research in
of organizational life (Oswick et al. 2004). Analyses focus 22.6 analysed the rhetorical dynamics and met-
often focus on their role in communicating with large aphors used by the charismatic former CEO of Apple
492 22 Language in qualitative research

22.6 RESEARCH IN FOCUS

The rhetorical construction of charismatic
leadership
Heracleous and Klaering (2014) were interested in exploring the effect of rhetorical competence on charismatic
leadership and whether leaders changed their rhetorical discourse in different contexts. Their empirical study of
this question relied on an in-depth case study of Apple Inc. former CEO, the late Steve Jobs. A key aspect of
charismatic leadership is suggested to involve the use of metaphors to communicate with followers. Their
approach involved the analysis of three texts which were selected to represent temporal and contextual diversity
and to reflect diversity in the leader’s perceived ‘ethos’, or credibility in the situation:

1. a Securities and Exchange Commission (SEC) deposition by Steve Jobs given in 2008 concerning stock option
backdating (119-page document comprising over 18,000 words), a situation of low ethos because Apple was
being investigated for potential illegal practices;
2. a CNCB interview with Steve jobs regarding Apple’s supplier shift from IBM to Intel conducted in 2005 (521
words), a situation of medium ethos where the leader was not particularly admired;
3. a discussion with Steve Jobs about media and technology in the Wall Street Journal, 2010 (transcript length:
12,006 words), a situation of high ethos where the leader was being ‘worshipped’.

One 500-word portion was selected for detailed analysis from each of the three texts, looking for central themes
and the root metaphors employed. In addition to ethos, or speaker credibility, the researcher focused on two other
aspects of rhetoric: ‘logos’, use of logic in arguments, and ‘pathos’, ability to ignite audience emotions.

Inc., the late Steve Jobs, in his communications with could be … And I felt that the Board wasn’t really do-
diverse audiences. ing the same with me … So I was hurt, I suppose would
This study highlights the importance of rhetorical be the most accurate word … I had been working, you
devices in provoking identification and commitment know, 4 years, 5 years of my life and not seeing my fam-
among listeners. It suggests that how a leader’s mes- ily very much and stuff and I just felt like there is nobody
sage is framed, through the use of metaphors, rhythm, looking out for me here …
contrasts, and lists, is as important as what the speech (Heracleous and Klaering 2014: 141)
is about in gaining commitment from followers. These
‘tools for framing’ define the form and construction By describing the Board of Director’s lack of care for him,
of the message by providing vivid images for the audi- Steve Jobs invokes the rhetorical dynamic of pathos to
ence. According to Heracleous and Klaering (2014), they evoke audience sympathy by portraying himself as ‘a self-
include pathos and metaphor. sacrificing businessman who places the company above his
own interests’ (Heracleous and Klaering 2014: 142). How-
• Pathos. Here the speaker, Steve Jobs, attempts to evoke
ever, in the interview with the Wall Street Journal, where
sympathy from the audience by presenting himself as
Steve Jobs is being interviewed as a respected expert and
a human being rather than as the CEO of a successful
invited to share his wisdom, he draws on a much wider
multi-million dollar company, as the following extract
range of rhetorical techniques, including metaphor.
from the SEC deposition illustrates:
• Metaphor. Steve Jobs also uses root metaphors, such as
Steve Jobs: Well it was a tough situation, you know. It the ‘circle of life’, as illustrated by the following extract:
wasn’t so much about the money … And as we’ve seen
in the discussions of the past hour, I spent a lot of time Steve Jobs: The way we’ve succeeded is by choosing
trying to take care of people at Apple and to, you know, what horses to ride really carefully, technically. We try
surprise and delight them with what a career at Apple to look for these technical vectors, that have a future
 Conversation analysis 493
and that are headed up and you know. Technology, dif- the four seasons to describe products, with ‘spring’
ferent pieces of technology, kinda go in cycles, they have referring to their birth, and ‘graveyard’ to describe their
their springs and summers and autumns, and then they, death. This ‘circle of life’ metaphor is also repeated in
you know, go to the graveyard of technology. So we try relation to the Apple company, which Jobs describes as
to pick things that are in their springs. ‘on its way to oblivion’, and then struggling for survival,
Here, Steve Jobs is speaking in a context of high ethos, before experiencing a rebirth.
i.e. he has strong credibility and therefore he adopts a Rhetorical analysis thereby enables a focus on the per-
more entertaining and expansive rhetorical style, using suasive acts that help to engender identification and fos-
ter co-operation within a group.

Conversation analysis
Conversation analysis is a fine-grained approach to the ethnomethodology satisfies two of the preoccupations of
analysis of language in use, whether in conversation or in qualitative researchers—the preference for a contextual
dialogue, that seeks to understand its organizing proper- understanding of action (see Chapter 17) and an ontolog-
ties—in other words, the rules and structures that deter- ical position associated with constructionism (discussed in
mine what people say in a given interaction. The roots of Chapter 2). The preference for analysing talk in naturally
conversation analysis lie in ethnomethodology, a sociologi- occurring situations suggests that conversation analysis
cal position developed by Harold Garfinkel and Harvey chimes with another preoccupation among qualitative
Sacks, though it is the latter with whom conversation researchers—namely, a commitment to naturalism.
analysis is most associated. Ethnomethodology is con- Two ideas are central to ethnomethodology and find
cerned with the ‘methods or procedures that competent clear expression in conversation analysis: indexicality
members of that social group use to go about “organiz- and reflexivity. Indexicality means that the meaning of
ing” themselves’ (Whittle 2018: 217). Ethnomethod- an act, which in conversation analysis essentially means
ologists are fundamentally concerned with the notion of spoken words or utterances including pauses and sounds,
organization as something that is worked at and accom- depends upon the context in which it is used. Reflexivity
plished through social processes of interaction, rather means that spoken words are constitutive of the social
than pre-given. Contrary to what its name implies, eth- world in which they are located; in other words, the prin-
nomethodology is not a research methodology; it is the ciple of reflexivity in ethnomethodology means that talk
study of the methods employed in everyday life though is not a ‘mere’ representation of the social world—it does
which social order is accomplished. much more than just stand for something else. While
The research that ethnomethodologists do involves ethnomethodological research is not a unified field, the
‘studying actual scenes as they unfold in real time in approach on which we focus here is associated with the
order to identify the kinds of knowledge and reason- work of Sacks (e.g. Sacks et al. 1974). It involves con-
ing—the “ethno-methods”—that enable people to ducting fine-grained analyses of talk in naturally occur-
organize themselves to accomplish some kind of joint ring situations and is referred to as conversation analysis
activity’ (Whittle 2018: 219). These ideas mean that (Key concept 22.7).

22.7 KEY CONCEPT

What is conversation analysis?
Conversation analysis is the fine-grained analysis of talk in naturally occurring situations. To conduct such
investigations, a premium is placed on the recording and transcription of naturally occurring conversations for the
purposes of intensive analysis. These analyses are concerned with uncovering the underlying structures of talk
and understanding how order is achieved through interaction.
494 22 Language in qualitative research

Heritage (1984, 1987) has proposed that conversation • Brackets indicate the point at which simultaneous
analysis is governed by three basic assumptions. speech overlaps—for example when more than one
speaker talks at the same time.
• Talk is structured. Talk comprises invariant patterns—
that is, it is structured. Participants are implicitly • A colon in the middle of a word indicates that the
aware of the rules that underpin these patterns. As a sound that occurs directly before the colon is pro-
result, conversation analysts eschew attempts to infer longed (e.g. we:ll). More than one colon means fur-
the motivations of speakers from what they say or to ther prolongation (e.g. ::::).
ascribe their talk to personal characteristics. Such
• The notation .hhh preceded by a dot indicate an
information is unnecessary, since the conversation intake of breath. If no dot is present, it means breath-
analyst is orientated to the underlying structures of ing out.
action, as revealed in talk.
• A bracketed full stop (.) indicates a very slight pause.
• Talk is forged contextually. Action is revealed in talk,
and as such talk must be analysed in terms of its con- The attention to detail in conversation analysis is very
text. This means that we must seek to understand what striking and represents a clear difference from the way
someone says in terms of the talk that has preceded it in which talk is normally treated by qualitative research-
and that therefore talk is viewed as exhibiting pat- ers such as when transcribing and analysing interviews.
terned sequences. Attention to fine details is an essential ingredient of con-
versation analysis work. Pauses and emphases are not to
• Analysis is grounded in data. Conversation analysts
be regarded as incidental or of little significance in terms
shun prior theoretical schemes and instead argue that
of what the speaker is trying to achieve; instead, they are
characteristics of talk and of the constitutive nature of
part of ‘the specific details of interaction [that] cannot
social order in each empirical instance must be
simply be ignored as insignificant’, as Heritage (1987:
induced out of data. Heritage (1987: 258) has written:
248) puts it. The gradual accumulation of detailed anal-
‘It is assumed that social actions work in detail and
yses of talk in interaction has resulted in recognition of
hence that the specific details of interaction cannot
recurring features in the way that talk is organized. These
simply be ignored as insignificant without damaging
features can be regarded as tools that can be applied to
the prospects for coherent and effective analyses.’ This
sequences of conversation.
assumption represents a manifesto for the emphasis
One of the most basic ideas in conversation analysis is
on fine-grained details (including length of pauses,
the notion that one of the ways in which order is achieved
prolongation of sounds, and so on) that is the hallmark
in everyday conversation is through turn-taking. This is
of conversation analysis.
a particularly important tool of conversation analysis,
As the third of the three assumptions indicates, conver- because it illustrates that talk depends on shared codes.
sation analysis requires the analyst to produce detailed If such codes did not exist, there would not be smooth
transcripts of natural conversation that includes all the transitions in conversation. In other words, there must be
pauses, interruptions, and intonations used by speakers. codes to indicate the ends of utterances. One of the ways
Some of the basic notational symbols employed in con- in which turn-taking is revealed is through the exami-
versation analysis are listed below. nation of adjacency pairs. The idea of the adjacency pair
draws attention to the well-attested tendency for some
• A figure in parentheses is used to indicate the length of
kinds of activity as revealed in talk to involve two linked
a period of silence, usually measured in tenths of one
phases: a question followed by an answer; an invitation
second. Thus, (0.3) signals three-tenths of a second of
followed by a response (accept/decline); or a greeting
silence.
followed by a returned greeting. The first phase invari-
• Punctuation marks, such as an exclamation mark, are ably implies that the other part of the adjacency pair will
used to capture characteristics of speech delivery be forthcoming—for example, that an invitation will be
rather than as grammatical notation. responded to. The second phase is of interest to the con-
• Italics are indicative of an emphasis in the speaker’s versation analyst not just because it becomes a spring-
delivery of a word. board for a response but because compliance with the
normative structure of the pairing indicates an appre-
• A hyphen represents a cut-off of a prior word or sylla-
ble, which may arise because a speaker is interrupted ciation of how one is supposed to respond to the initial
by someone else. phase. In this way, ‘intersubjective understandings’ are
 Conversation analysis 495
continuously reinforced (Heritage 1987: 259–60). This The insistence of conversation analysts that understand-
is not to imply that the second phase will always follow ing must be based on sequences of talk, and must avoid
the first; indeed, the response to a failure to comply with extraneous inferences about the meanings of that talk,
the expected response is itself the focus of attention by marks it as a somewhat different approach from much
conversation analysts. qualitative research. As we have seen in previous chap-
Conversation analysts study talk in a range of orga- ters, qualitative researchers often seek to achieve under-
nizational settings, such as television news interviews, standing from the perspective of those being studied.
courtroom trials, and medical settings. For example, Conversation analysts claim to do this only in so far as
Boden (1994) uses conversation analysis to explore how that understanding can be revealed in the specific con-
talk is organized in formal and informal meetings. She texts of talk. To import elements that are not specifically
shows how interactional order is constructed through grounded in the here-and-now of what has just been said
sequences of talk that enable people to transmit informa- during a conversation risks implanting understanding
tion, make decisions, and sort out misunderstandings. It that is not grounded in participants’ own terms (Sche-
has sometimes been suggested that conversation analysis gloff 1997). In so doing, conversation analysis reduces
fails to capture body movements. However, the growth the risk of making unwarranted speculations about what
of video-based methods has enabled this limitation to is happening in social interactions and has contributed
be addressed, as the analysis of teamwork in hospital much to our understanding of the accomplishment of
anaesthesia teams by Hindmarsh and Pilnick illustrates social order.
(Research in focus 22.8).

22.8 RESEARCH IN FOCUS

A study of hospital teamwork using ethnomethodol-
ogy and conversation analysis
Hindmarsh and Pilnick’s (2007) research focused on understanding embodiment in the workplace. The
organizational setting that they selected for study was a hospital pre-operative anaesthetic team, where
embodied conduct and coordination is critical in ensuring successful outcomes. They describe their approach
as involving consideration of ‘the fine-grained, closely organized practices of teamwork among personnel
involved in administering and managing anaesthetic procedures’ (2007: 1396). In addition to interviewing
members of the team and conducting observational fieldwork, a key method of data collection involved
‘naturalistic audio-visual recordings of everyday anaesthetic work’ (2007: 1398). These recordings were made
on 14 days over a period of several months. Crucially, this method enabled the study of non-vocal, i.e. embodied
interactions, in addition to talk.

A key analytical focus was on the ‘sequential’ organization of activity in ‘real-time’ as a means of understanding
how each ‘participant’s conduct emerges in relation to the actions of others’ (2007: 1401). They present their
data using a series of ‘fragments’, very short (around 15 seconds) extracts from the video recording. These are
presented in conjunction with a sequence of still images showing how the bodies of organizational members
move as they work. Plate 22.1 shows an extract from an image sequence involving two members of the
anaesthetic team conducting a procedure on a patient. What is striking about this example is relative absence of
talk, for as the researchers’ note ‘aside from a quiet “okay” by the anaesthetist at one point, this is all completed
without talk’ (2007: 1404). By presenting and analysing these short sequences of everyday anaesthetic work,
the researchers demonstrate the competence of team members as evidenced by their ability to sensitively and
seamlessly coordinate their bodily movements, a practice which they refer to as ‘intercorporeal knowing’. As this
example demonstrates, conversation analysis has moved beyond analysis of talk in isolation towards
incorporation of the non-verbal aspects through which interactional orders are produced.