Definition
Discourse analysis is the study of language in context. It is an approach that emphasizes how versions of the social world are produced within naturally occurring spoken or written discourse. The discourse analytic view is that all features of talk or texts perform some kind of action (for example exercizing power and control over others) and it is possible to analyse how language is used to achieve that action. It is therefore concerned with how participants construct themselves and others through their discourse and how these selves may be undermined.
Distinctive Features
There are a number of different approaches to discourse analysis (see Phillips and Jørgensen, 2002 for a full review) due in part to the developing nature of the field and to the number of disciplines in which it has its roots, including linguistics, sociology, psychology, philosophy and literary theory. However, all approaches share the same postmodern perspective that talk does not neutrally reflect our world and our social relationships but rather actively constructs and alters them. Discourse theory views discourse as a constantly changing entity and explores the struggle between competing forms of dis- courses each representing a particular way of understanding the world. Critical
Discourse Analysis
53 Key Readings
*Elliot, H. (1997) ‘The use of diaries in social research’, Sociological Research
Online, 2(2). http://www.socresonline.
org.uk/socresonline/2/2/4.html Holliday, R. (2004) ‘Filming “the closet”:
the role of video diaries in researching sexuality’, American Behavioral Scien-
tist, 47(12): 1597–1616.
*Johnson, J. and Bytheway, B. (2001) ‘An evaluation of the use of diaries in a study of medication in later life’,
International Journal of Social Research
Methodology: Theory & Practice, 4(3):
183–204.
Lee, R.M. (1993) Doing Research on
Sensitive Topics. London: Sage.
Office of National Statistics (1999)
Family Spending: A Report on the 1998– 1999 Family Expenditure Survey. London:
The Stationary Office.
Plummer, K. (1983) Documents of Life. London: Allen & Unwin.
Zimmerman, D.H. and Wieder, D.L. (1977) ‘The diary-interview method’,
discourse analysis accepts that discourse is just one among many aspects of social practice and that intertextuality (that is, how texts draw on elements of other texts) is a central concern. Discursive psychology sees individual actors as both the products and the producers of discourse and is more concerned with small-scale ‘talk-in-interaction’ rather than large-scale societal discourse.
Discourse analysis is not just a method of data analysis but an approach that combines philosophy, theory and method. Discourse analysts start with the acceptance that discourse is a form of social action that plays a part in producing the social world (including social relations and knowledge). Physical objects and events exist independently of people’s thoughts and speech, but they have meaning only through discourse. For instance, most people would view cancer as an illness, but they would not necessarily describe it in the same way. Some would draw on discourses of behavioural risk, while others may see it as a result of medical mismanagement and others still might attribute it to God’s will. Importantly, each of these discourses of explanation will suggest a different course of action to tackle the illness.
Discourse analysis can be applied in a number of data collection settings primarily to examine naturally occurring talk. For example discourse analysts might be interested in talk within institutions (schools, hospitals, prisons) or interested in written texts such as newspapers. Discourse analysis might be applied to examine a number of subjects such as gender inequalities, national identity or the construction of knowledge claims. Questions are not just asked about the content of the discourse but also about the author, its authority and its audience. Consequently discourse analysis is often applied in conjunction with documentary methods.
Discourse analysis also has similarities with conversation analysis, not least because of its use of naturally occurring speech. However, it differs from conversation analysis in that it deals with wider social science concerns such as gender relations and social control. Whereas conversation analysts exclude the context in which participants speak, discourse analysts argue that one cannot understand what is going on in a particular interactional episode unless one knows how to locate it within the macro societal context. Discourse analysis takes meaning above the utterance level to focus on the participants’ roles and their institutional or ideological motives. Therefore, although audio-recordings and transcription are required to be of good quality, discourse analysis does not require the same attention to detail as the units of analysis are broader.
Examples
Sarangi et al. (2003) present a discourse analysis of data derived from genetic counselling clinics. Focusing on the discourse of risk assessment and risk
communication they identify six strategies – abstraction, reformulation, externalization, localization, temporalization and agentivization – by which both patients and counsellors seek to relativize risk status. For example the strategy of abstraction involves statements of the risks of other individuals: ‘about one in three people get a cancer somewhere in their body in their life’. In contrast the strategy of reformulation places risk in the very personal context of the individual’s life. It is often coded in figures of speech and metaphors in order to be made relevant to the patient’s lived experiences: ‘… it’s like betting on a horse, and so it’s important for you not to say oh well it’s only one in seven, couldn’t happen to me. It’s important to think through the issues.’ The authors conclude that it is through the use of these strategies that counsellors are able to present the same risk information in either more reas- suring or more alarming terms, thereby balancing the needs of accurate transmission of facts with promoting appropriate behaviour and avoidance of unnecessary anxiety.
An example of a discourse analysis of textual material is Torck’s (2001) cross-cultural study of the voices of homeless people in street newspapers. Torck’s analysis challenges the claim that the purpose of street newspapers is to raise the profile of homelessness issues and to be a platform for homeless people to regain independence. The article shows how papers’ topics and genres are framed to reinforce the negative social ethos of the homeless. For example where homeless people are given a platform for their voice they are usually limited to personal narratives and poetry. The emphasis on feelings and pathos in the voices of the homeless was found to be less prevalent in the American compared to the European papers.
The idea that discourse constructs the world rather than representing it has also been popular within the sociology of science. Research has focused on the importance of scientific discourse and texts particularly in relation to sci- entific persuasion and practical reasoning. For example Mulkay and Gilbert (1982) show how scientists use two distinct interpretative repertoires, or lin- guistic registers, when discussing their work. The empiricist repertoire is used by scientists when discussing work in a formal context, and is characterized by a conventionally impersonal style, where references to human actors are min- imized and the natural world appears to speak for itself. In contrast the con- tingent repertoire is characterized by references to the personal and social contingencies in scientific action and belief. Mulkay and Gilbert (1982) describe how scientists apply these two repertoires asymmetrically to account for correct and incorrect beliefs. Thus they present correct belief, which is invariably identical to their current views, as arising unproblematically from the experimental evidence, and incorrect belief is explained by reference to the distorting effects of personal and social factors.
Discourse Analysis
Evaluation
Discourse analysis has been described as a craft skill (Potter, 1997). Researchers are required to develop an analytic mentality to their data that comes with experience, although familiarity with other discourse analytic studies will aid the grasp of general principles and methodological strategies. Wood and Kroger (2000) provide a number of suggestions and analytical concepts for novice dis- course analysts in order to develop their analytic resources and assist with gen- erating interpretations of the data. These suggestions include considering what is missing from the discourse (for example, an apology or a greeting) and con- sidering how the discourse makes the reader feel (for example, amused or angry), and trying to identify the features of the text that produced such feelings.
There are concerns that discourse might be affected by the act of audio- recording, in which case the data cannot be said to be naturally occurring. There are occasions when recording would not result in bias, for example when recording from a radio broadcast. Even if speakers are aware that their speech is being recorded, the discourse might not be affected if, for example, they are highly involved with the task at hand or if recording takes place over periods of time (Wood and Kroger, 2000). Discourse analysis has been criticized for not attending to the non-verbal aspects of interaction. Such concerns emerge from a belief that non-verbal interaction is more trustworthy than verbal because it is less controllable and hence more likely to disclose true meaning. This position fails to recognize that language is an action and indeed that most people seem to have difficulty con- trolling the details of their speech. In fact discourse analysts do attend to the non-verbal. Potter’s (1997) analysis of the BBC interview with Princess Diana explicitly draws attention to the significance of the Princess’s non-verbal actions (head shaking and distant gazes) within the exchange.
Discourse analysis is a generic term covering a range of approaches and perhaps would not be considered by some as a method at all. However, discourse analysis has been influential in enabling researchers to expose the inconsis- tencies and inequities within social relationships.
Associated Concepts:
Audio-Recording, Bias, Conversation Analysis,
Documentary Methods, Transcription.
56
Key Readings
Cameron, D. (2001) Working with Spoken
Discourse. London: Sage.
Mulkay, M. and Gilbert, G.N. (1982) ‘Accounting for error: how scientists construct their social world and
how they account for correct and incorrect belief’, Sociology, 16(2): 165–183.
Phillips, L. and Jørgensen, M. (2002)
Discourse Analysis as Theory and Method.