Chapter 23: Discourse analysis
Tess Tsindos
Learning outcomes
Upon completion of this chapter, you should be able to:
- Describe discourse analysis.
- Understand how to conduct discourse analysis.
- Identify the strengths and limitations of discourse analysis.
What is discourse analysis?
Discourse analysis is a field of qualitative analysis that has its origins in disciplines such as linguistics, philosophy, psychology, anthropology.1 It is an interdisciplinary field that deals with ‘language’ and meaning.2
According to Jaworski and Coupland, the purpose of discourse analysis is that it ‘offers a means of exposing or deconstructing the social practices that constitute ‘social structure’ and what we might call the conventional meaning structures of social life. It is a sort of forensic activity’.3(p5) There are three domains of discourse analysis: the study of social interaction; the study of minds, selves and sense-making; and the study of culture and social relations.4(p5)
Discourse analysis is the study of texts such as transcribed interviews, websites, forums, books, newspapers, government documents (and many more), and the analysis of those texts to understand different accounts and the meanings behind those accounts. Qualitative researchers strive to understand the relationships between text (discourse) and social constructs. As text is analysed, the meaning behind the text is also explored, often as the ‘voices’ in the text. For example, when a participant is asked about their eating habits and they discuss their joy in eating as well as feelings of guilt from eating high-calorific foods, they may be voicing their parents’ disapproval of this eating behaviour. The relationship between text and social constructs can also be seen in alcohol advertising: an advertisement may be promoting alcohol consumption as a fun behaviour, but also cautions listeners to drink ‘responsibly’, because the advertiser is required to do so by advertising standards authorities. This inherent contradiction in the advertising is part of the meaning-making regarding alcohol consumption. This meaning-making is contextual and differs between countries, such as Australia (a high alcohol consumption culture) and Canada (a lower alcohol consumption culture). Another example of context is in the use of the word ‘just’ by an interview participant; the term can mean many things, but if the researcher is asking about job title, ‘just’ may the participant’s implication or inference that the title does not reflect an important position (e.g. ‘I’m just an editor’). In discourse analysis, texts, meanings and inferences are important.
Following is an example of media articles and two distinct discourses about violence towards women. The first media article, published by The Guardian on 15 June 2018,5 presents a discourse about how it is the responsibility of women to prevent men from being violent towards them. The second article about the same incident, published by The Age on 25 May 2019,6 presents a discourse that it is the responsibility of men not to be violent towards women.
Meanings of texts are particularly important when participants use metaphors. The researcher needs to examine the implications of the metaphor, deliberate or inadvertent. For example, when the researcher asks the participant how they felt about their life and the participant replies, ‘life is a highway’, the researcher needs to look beyond what was said to understand the participant’s meaning.
As an interdisciplinary method, discourse analysis can be complex and intricate. Gee7 provides 72 tools to assist with various types of discourse analysis, ranging from identifying what is being said and what is not being said, to examining ‘how the person is using language, as well as ways of acting, interacting, believing, valuing, dressing, and using various objects, tools, and technologies in certain sorts of environments to enact a specific socially recognizable identity and engage in one or more socially recognizable activities’.7(p201) Gee also includes a helpful table (see Table 23.1) populated with his 7 building tasks for researchers to examine their discourses, and provides the answers.8
Table 23.1. Seven Building Tasks and associated discourse analysis questions
| Gee’s building task | Gee’s discourse analysis question |
|---|---|
| Significance | How is this piece of language being used to make certain things significant or not, and in what ways? |
| Practices | What practice or practices is this piece of language being used to enact (i.e. to get others to recognise as going on)? |
| Identities | What identity or identities is this piece of language being used to enact (i.e. to get others to recognise as operative)?What identity or identities is this piece of language attributing to others, and how does this help the speaker or writer enact his or her own identity? |
| Relationships | What sort of relationship or relationships is this piece of language seeking to enact with others? |
| Politics | What perspective on social goods is this piece of language communicating (i.e. what is being communicated as to what is taken to be ‘normal,’ ‘right,’ ‘good,’ ‘correct,’ ‘proper,’ ‘appropriate,’ ‘valuable,’ ‘the ways things are,’ ‘the way things ought to be,’ ‘high status or low status,’ ‘like me or not like me,’ and so forth)? |
| Connections | How does this piece of language connect or disconnect things; how does it make one thing relevant or irrelevant to another? |
| Sign systems and knowledge | How does this piece of language privilege or deprivilege specific sign systems or different ways of knowing and believing, or claims to knowledge and beliefs? |
How to conduct discourse analysis
Discourse analysis, as in all other qualitative methods, is used depending on the research topic and question(s) or aim(s). The following steps are recommended:
Step 1: Have a clearly defined topic and research question, because this informs the types of research materials that will be used.
Step 2: Conduct wide-ranging searches for materials that will inform the research topic.
Step 3: Determine which theory and framework will be used as the underpinning foundation for the analyses (see Section 1 chapters 1–4).
Step 4: Analyse the content of the materials. This analysis is different (but similar) to content analysis, which is a research technique to systematically classify codes and identify themes or patterns within the data. Discourse analysis is concerned with identifying themes and patterns within the texts that relate to the social contexts reflected in the research topic and within the theoretical lens chosen for analyses.
Step 5: Interpret and draw conclusions. Reflect on your work and examine how the various texts use language within the context of the research topic to answer the research question(s).
As an example, Table 23.3 includes a study on girls’ experience of competitive dancing.9 The authors progressed through the steps as follows:
Step 1: The topic is eating disorders and young dancers. The research question is ‘How does experience in the world of competitive dance shape the relationship that young girls have with their bodies?’
Step 2: The author conducted wide–ranging literature searches on eating disorders, ballet dancers, body image, thinness, Western culture, dieting, media influences and many more topics.
Step 3: Feminism was the theoretical underpinning of the textual analysis. As described by the authors, ‘a feminist post structural approach was chosen to provide a critical lens to explore the beliefs, values, and practices of young dancers… aimed to provide an understanding of the dominant and competing discourses present in the world of dance and discover how these discourses are constituted, perpetuated, and form ways of knowing in relation to body and body image.’9(p7)
Step 4: The transcripts were analysed in 5 steps, following Aston10 and presented in table 23.2:
Table 23.2. A guide to using feminist poststructuralism informed by discourse analysis
| 1. Identify important issues | Read the transcript and mark quotations you feel represent an important issue. Name the issue as you see it. |
| 2. Apply beliefs, values and practices | Provide the quotation (cut and paste) and write something about the belief, value and practice within the quotation. |
| 3. Social and institutional discourses | Write about the social and institutional discourses you see informing the issue you identified. Sometimes this is clearly described in the quotation but most often you will need to expand on the implied ideas. You still need to clearly connect to the evidence (words and meaning provided by participant). |
| 4. Respond to relations of power | As you write about the discourses, you need to connect these ideas to the participant.How do the discourses affect the participant? Does he/she agree or disagree with the beliefs, values and practices? Is it an easy or positive fit? Or are there questions, conflicts, tensions etc.? These are the ‘relations of power’ that the participant is feeling experiencing. |
| 5. Subjectivity and agency | You can also add in the participant’s ‘subjectivity’ (how they are positioned as a nurse, man, woman, teacher etc.) as well as their ‘agency’ (how they choose to act in each situation by fitting in or challenging). |
*Note: This table is from an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits copy and redistribution of material in any medium or format, remix, transform and build upon the material for any purpose, even commercially provided the original work is properly cited.
Step 5: Results were first interpreted within an ‘environmental’ context (competitive culture, ideal dancer’s body, mirrors, and dance attire and costumes), which was predominately negative due to the competitive culture. The second context was ‘parents’, which encompassed body monitoring, joking, and parents and support. Although most of the dancers stated that their parents did not influence their relationship with their body, discourse analysis demonstrated that parents did influence them. The third context was ‘coaches’. Coaches had a very strong influence on participants’ body image. While the dancers believed their coaches were supportive, the discourse demonstrated that most coaches’ comments were negative. ‘Peers’ represented in the final context for analysis. Again, the dancers believed their peers were supportive; however, discourse analysis demonstrated that many peer comments were negative. The conclusions drawn from the research were that ‘all participants experienced negative physical, mental, and/or emotional repercussions throughout their competitive dance experience. It was also determined that environment, parents, coaches, and peers largely shaped the dancer’s relationship with body and body image in the world of dance. These influences generated and perpetuated the dominant negative body image discourse that dancers were often unable to resist, and consequently their relationship with body and body image suffered.’9(pp22-23)
This is a good example of situating a topic (body image) within a context (young women dancing) underpinned by a theoretical framework that explores the dancers’ beliefs, values and practices.
Table 23.3. Discourse analysis examples
| Title | The public health turn on violence against women’: analysing Swedish healthcare law, public health and gender-equality policies11 | Social representation of palliative care in the Spanish printed media: a qualitative analysis12 | Dancing in a culture of disordered eating: a feminist post structural analysis of body and body image among young girls in the world of dance9 |
|---|---|---|---|
| CC Licence | CC BY 4.0 | CC BY 4.0 | CC BY 4.0 |
| First author and year | Ohman, 2020 | Carrasco, 2019 | Doria, 2022 |
| Aim/research question | ‘To describe and problematise the main content and characteristics of Swedish healthcare law, public health and gender-equality policies representing the public health turn on violence against women.’(p.2) | ‘To analyse how palliative care is portrayed in Spanish newspapers, as well as the contribution made by the press to its social representation.’ (abstract) | 'How does experience in the world of competitive dance shape the relationship that young girls have with their bodies'(p.7) |
| Study design | Multidisciplinary, socio-legal | Qualitative | Qualitative |
| Data collection | National healthcare law and policies | Four Spanish general printed newspapers | One-on-one, semi-structured phone interviews, directed by an open-ended interview guide |
| Analysis approach | Discourse analysis | Discourse analysis | Discourse analysis |
| Theoretical approach | Legal documents primarily analysed from a feminist legal point of view; public health actions and interventions analysed from a public health perspective; and general gender-equality policies analysed from a policy angle | Sociological discourse analysis: contextual analysis focusing on the message as a statement; interpretative analysis considering the discourse as a social product | Feminist poststructuralismp |
| Results | In law and public health policies, the problem is primarily articulated as a matter of ‘violence within close relationships' The term ‘violence within close relationships’ is a new approach that deviates from the earlier framings of ‘men’s violence against women’, and is a specific Swedish policy term. This new approach indicates a gender-neutral conceptualisation in which both victim and perpetrator are invisible in terms of gender.Legal obligations and the problems for the healthcare sector are only vaguely defined. |
‘The discourses identified were characterised by strong ideological and moral content focusing on social debate, strong ties linking palliative care and death and, to a lesser degree, as a healthcare service.The messages transmitted by representatives with direct experience in palliative care (professionals, patients and families) contributed the most to building a positive image of this healthcare practice. Overall, media reflect different interests in framing public understanding about palliative care.’ (abstract) | 'All participants experienced negative physical, mental, and/or emotional repercussions throughout their competitive dance experience. It was also determined that environment, parents, coaches and peers largely shaped the dancers’ relationship with body and body image in the world of dance. These influences generated and perpetuated the dominant negative body image discourse that dancers were often unable to resist, and consequently their relationship with body and body image suffered.' (p.23-24) |
Advantages and challenges of discourse analysis
Discourse analysis can be used to analyse small and large data sets with homogenous and heterogenous samples. It can be applied to any type of data source, from interviews and focus groups to diary entries, news reports and online discussion forums. However, interpretation in discourse analysis can lead to limitations and challenges that tend to occur when discourse analysis is misapplied or done poorly. Discourse analysis can be highly flexible and is best used when anchored in a theoretical approach. Because discourse analysis involves subjective interpretation, training and support from a qualitative researcher with expertise in the method is required to ensure that the interpretation of the data is meaningful. Finally, discourse analysis can be time-consuming when analysing large volumes of texts.
Summary
Discourse analysis is a process whereby texts are examined and interpreted. It looks for the meanings ‘behind’ text in cultural and social contexts. Discourse analysis is flexible, and the researcher has scope to interpret the text(s) based on the research topic and aim(s). Having a theoretical approach assists the researcher to position the discourse in cultural and social grounding.
References
- Schiffrin D, Tannen D et al., eds. The Handbook of Discourse Analysis. Blackwell; 2001.
- Jaworski A, Coupland N. eds. The Discourse Reader. 2nd ed. Routledge; 2006.
- Jaworski A, Coupland N. Introduction: perspectives on discourse analysis. In: Jaworski A, Coupland N, eds. The Discourse Reader. 2nd ed. Routledge; 2006.
- Wetherell M, Taylor S, Yates S. (2001) Discourse Theory and Practice: A Reader. 2nd ed. Sage. 2001.
- Davey M. ‘Men need to change’: anger grows over police response to Eurydice Dixon’s murder. Guardian. June 15, 2018. Accessed April 28, 2023. https://www.theguardian.com/australia-news/2018/jun/15/men-need-to-change-anger-grows-over-police-response-to-comedians#:~:text=Melbourne
- Fowler M. ‘This is about men’s behaviour’, says top policy offer after another woman’s murder. Age. May 25, 2019. Accessed April 28, 2023. https://www.theage.com.au/national/victoria/this-is-about-men-s-behaviour-says-top-police-officer-after-another-woman-s-murder-20190525-p51r46.html
- Gee J. How to do Discourse Analysis: A Toolkit. 2nd ed. Routledge; 2014.
- Gee J. An Introduction to Discourse Analysis: Theory and Method. 3rd ed. Routledge; 2011.
- Doria N, Numer M. Dancing in a culture of disordered eating: a feminist poststructural analysis of body and body image among young girls in the world of dance. PLoS ONE. 2022;17(1): e0247651. doi:10.1371/journal.pone.0247651
- Aston M. Teaching feminist poststructuralism: founding scholars still relevant today. Creative Education. 2016;7(15):2251-2267. doi:10.4236/ce.2016.715220
- Öhman A, Burman M, Carbin Met al.‘The public health turn on violence against women’: analysing Swedish healthcare law, public health and gender-equality policies. BMC Public Health. 2020;20:753. doi:10.1186/s12889-020-08766-7
- Carrasco JM, Gómez-Baceiredo B, Navas A et al. Social representation of palliative care in the Spanish printed media: a qualitative analysis. PLoS ONE. 2019;14(1):e0211106. doi:10.1371/journal.pone.0211106
