Discourse analysis and discursive psychology

Individual workYSLU

By

Maryam Eskandarjouy

Professor G. Hovhannisyan

First year M.A student

29 June 2013

Definition

Discourse – defined as language use in everyday text and talk – is a dynamic form of social

practice which constructs the social world, individual selves and identity. The self is

constructed through the internalization of social dialogues. People have several, flexible

identities which are constructed on the basis of different discourses. Power functions

through the individual’s positioning in particular discursive categories. Discourse does not

give expression to pre-constituted psychological states; rather, subjective psychological

realities are constituted in discourse. Individuals’ claims about psychological states should

be treated as social, discursive activities instead of as expressions of deeper ‘essences’

behind the words.

• Discourse is best viewed not as an abstract system (the tendency in structuralism and

poststructuralist theories of discourse) but as ‘situated’ language use in the contexts in which

it takes place.

• People use discourse rhetorically in order to accomplish forms of social action in particular

contexts of interaction. Language use is, in this sense, ‘occasioned’. The focus of analysis,

then, is not on the linguistic organization of text and talk as in critical discourse analysis but

on the rhetorical organization of text and talk. The following questions are asked. What do

people do with their text and talk?

How are accounts established as solid, real and stable representations of the world? How are

people’s constructions of the world designed so that they appear as stable facts, and how do

they undermine alternative versions (‘dilemmas of stake’)?

• Language constitutes the unconscious as well as consciousness. Psychoanalytical

theory can be combined with discourse analysis in order to account for the

psychological mechanisms underpinning the ‘unsaid’ and people’s selective

investment in particular discourses from the range of available discourses.

• The understanding of the contingent nature of research knowledge leads to

reflexive consideration of issues relating to relativism and the role of the

researcher in knowledge production.

Different strands of discursive psychology

In short, the three strands can be described as follows:

• A poststructuralist perspective that builds on a theory on discourse, power and

the subject.

• An interactionist perspective that builds on conversation analysis and ethno

methodology.

• A synthetic perspective that unites the two first perspectives.

The differences between the three strands can be illustrated by way of the continuum.

On the right-hand side lie the approaches in which the researcher identifies abstract

discourses without examining in detail their use across different social contexts. On

the left-hand side are the approaches in which the researcher investigates details in

language use as activities in social interaction without systematically analyzing the

links between the details and broader social and cultural processes and structures. The

first perspective belongs to the right-hand side, the second perspective to the left-hand

side, and the third to the middle position. The focus in the first perspective, closest to

the more abstract conception of discourse, then, is on how people’s understandings of

the world and identities are created and changed in specific discourses and on the

social consequences of these discursive constructions.

The second perspective concentrates on the analysis of the action orientation of

text and talk in social interaction. Drawing on conversation analysis and ethno

methodology, the focus is on how social organization is produced through

speech and interaction. The researcher analyses people’s conversations as

manifestations of a world that the participants create themselves. The aim of

the researcher is to keep her/his own theoretical perspective on this world out

of the analysis, and it is considered to be an assault on the empirical material to

apply frames of understanding and explanation not thematised by the

informants’ themselves.

In the third perspective, a poststructuralist interest in how specific discourses

constitute subjects and objects is combined with an interactionist interest in the

ways in which people’s discourse is oriented towards social action in specific

contexts of interaction.12 Equal stress is placed on what people do with their text

and talk and on the discursive resources they deploy in these practices. The

concept of interpretative repertoire is often used instead of discourse to emphasize

that discourses are drawn on in social interaction as flexible resources.

Proponents of this synthetic perspective distance themselves from both

poststructuralist discourse analysis and conversation analysis in their

unadulterated forms. On the one hand, they criticize poststructuralist discourse

analysis for reifying discourses – treating them as things out there in the world –

and for neglecting people’s situated language use (for example, Wetherell 1998).

In poststructuralist discourse analyses of a particular domain (such as the domain

of sexuality, politics or the media), it is argued, discourses are viewed as

monolithic structures to which people are subjected, and insufficient account is

taken of the ways in which people’s talk is shaped and changed by the specific

contexts of interaction in which the talk is situated and to which it is oriented.

On the other hand, they argue that conversation analysis as practiced both in the

field of conversation analysis itself and in the purely interactionist perspective

in discursive psychology14 neglects the wider social and ideological

consequences of language use (for example, Billig 1999a, b;15 Wetherell 1998).

These consequences, it is proposed by followers of the synthetic perspective

such as Wetherell (1998) and Billig (1999b), can – and should – be explored

through application of social theory in addition to conversation analysis or

discourse analysis.

This option is ruled out by conversation analysts on the grounds that the proper object of

analysis is the participants’ own meaning-making through talk-in-interaction and not the

analysts’ interpretations of that talk in terms of the wider social patterning of talk. But this

claim to produce an analysis of participants’ own understandings free from the ‘pollution’

of analytical assumptions is an expression of epistemological naivety as well as being

undesirable from the perspective of critical research, according to Billig (1999b). Although

we will refer to the first two strands in cases of disagreement between the three strands of

discursive psychology, in most of this chapter we concentrate mainly on the third

perspective, focusing on the work of Potter and Wetherell, since their approach has been

central for the development of discursive psychology in general and provides particularly

useful and widely used tools for research in communication, culture and language.

Discursive research treats discourse as having four key characteristics:

1. Discourse is action-oriented.

2. Discourse is situated.

3. Discourse is both constructed and constructive.

4. Discourse is produced as psychological.

Seven Stages in the Execution of Discursive Research

Stage One: Obtaining Access and Consent

One of the features that make contemporary discursive psychology

distinctive from most other psychological methods is that it works primarily

with audio or video records of interaction happening in natural settings.

This makes the process of gaining access and consent, developing

appropriate ethics scripts, and working closely with participants in a way

that sustains and merits a strong degree of trust an integral part of the

research process.

Gaining access and consent can be a challenge. And it is likely that researchers

sometimes use other forms of data generation – questionnaires, say, or open ended

interviews – because they expect that access will be refused. However, experience

shows that with the right approach and a proportionate commitment of time and

effort trust can be developed and consent can be obtained for working in the most

sensitive of sites. Initial contact is often through a key institutional member – a

medical practitioner, school teacher or parent – who can provide an authoritative

link for the researchers. A key feature of this contact is often to identify the

participants’ anxieties about the research process. These are often focused on the

possibility that the research will evaluate their practice.

Stage Two: Data Collection

In terms of data collection the main aim is to develop an archive of records of interaction in the setting

under study. There are no hard and fast rules for the size of such a collection. Even small amounts of

material can be a basis for useful research; but the more material there is and the more appropriate the

sampling then more questions will become analytically tractable and more confidence can be placed in

the research conclusions. Time and resources devoted to getting high quality recordings will pay off

handsomely when it comes to transcribing the recordings and working with them in data sessions. Solid

state recorders with good microphones and digital video cameras with large hard disc drives are both

effective. It will be important to have video records of face to face interaction. Researchers should err on

the side of collecting more recordings than planned. Digital recordings can be easily stored and they

provide an important resource for future research. Participants do the data collection themselves.

Simplicity is a key consideration – it minimizes what the participants have to learn and the effort they

have to put into the collection.

Stage Three: Data Management

Much of this is focused on systems of folders that collect together recordings in different forms,

different forms of transcript, and analytic notes. Such a system can facilitate data sharing and can assist

full backup of data and analysis. Coding and secure storage may be required depending on the

agreements with participants and the sensitivity of the materials. This is also a task for data reduction

and involves the systematic building of a particular corpus that is of a size small enough to be easily

worked with but large enough to be able to make appropriate generalizations. In the NSPCC study we

assigned a two letter code to each CPO who took part; each had their own folder. Within each folder

each call had its own folder with a memorable name. Within this folder was a high quality recording in

WAV format, but also a smaller MP3 version that could be emailed and easily backed up. Each folder

often also contained two versions of the transcript and sometimes further transcript and analytic

observations.

Stage Four: Transcription

Discourse research works continuously with both the original audio or video recordings and

the transcript. The transcript is an essential element in the research. It is common to use

two forms of transcript. A basic ‘first pass’ transcript is often generated by a transcription

service. This has just the words rendered as effectively as the service can hear them.

It allows the researcher to quickly go through a stretch of interaction and get an overall feel

for what is there and it is searchable, too.

The second form of transcription is an attempt to capture on the page features of the

delivery of talk that participants treat as relevant for understanding the activities that are

taking place. Because of the time investment in producing quality transcript there are rarely

resources for completely transcribing a full set of recordings. Various criteria can be used to

decide what to transcribe and in what order.

Stage Five: Developing Research Questions

It has been common in psychological research to stress the importance of formulating a clear research

question before starting the research. And there are often good reasons for such a rule as it can help

avoid confusion and sloppiness when doing a wide range of psychological studies, particularly when

utilizing experimental designs, questionnaires or open ended interviews. However, with discursive

research much of the discipline comes from working with a set of naturalistic materials – records of

people living their lives in a particular setting. And many of the questions formulated for more

traditional research have a causal form – what is the effect of X on Y – which is rarely appropriate for

discourse work. One of the benefits of working with naturalistic materials is that they throw up their

own challenges that lead to novel questions.

Stage Six: Analysis

In discursive psychology the analytic stage of work is often the most time consuming and

the most crucial. The different types of discursive psychology have different ways of

approaching discourse analysis. The choice of analytical techniques depends on the

theoretical frame and method. Discursive psychological analysis often uses a systematic

trawl through the materials to build a corpus of examples. When analytic understanding

has been improved it is likely that some of the cases will be dropped from the corpus and

new cases will be seen as appropriately part of the corpus.

Stage Seven: Validation

In practice there is not a clear-cut distinction between analysis and validation.

Building a successful analysis that works and is attentive to all the details of

the materials that are being studied is already a major part of validating the

findings.

Some themes:

1. Participants’ orientations. One of the enormous virtues of working with open ended,

naturally occurring materials is that they provide a major resource in validating findings that

is absent in most other psychological methods. That resource is the turn-by-turn nature of

interaction. Any turn of talk is oriented to what came before, and sets up an environment for

what comes next. Each turn provides, in its orientation to what came before, a display that is

central to the intelligibility of interaction. One of the limitations of most psychological

methods is that they cut across this kind of display.

2. Deviant cases. Studies of media interviews show that interviewees rarely treat

interviewers as accountable for views expressed in their questions.

3. Coherence. The accumulation of findings from different studies allows new

studies to be assessed for their coherence with what comes before.

4. Readers’ evaluation. One of the most fundamental features of discursive

psychology compared to other psychological perspectives is that its claims are

accountable to the detail of the empirical materials, and that the empirical

materials are presented in a form that allows readers, as far as possible, to make

their own checks and judgments.

Acknowledgment

I would like to acknowledge and

extend my heartfelt gratitude to

Professor G. Hovhannisyan for her

vital encouragement and support.

Most especially to my friends, And

to God, who made all things

possible.