Discourse Analysis: What It Is and How to Approach It

Discourse analysis is one of the more conceptually slippery qualitative research methods, partly because "discourse" itself is used in several related but distinct ways across different academic traditions. This guide sets out what discourse analysis actually studies, the main approaches you are likely to encounter, and practical guidance for applying it in your own dissertation.

What "Discourse" Actually Means Here

In everyday use, "discourse" just means conversation or communication. In the academic sense relevant to discourse analysis, discourse refers to language in use - not just individual sentences in isolation, but language as it functions within a specific social, institutional, or cultural context, and the ways that language both reflects and actively shapes how people understand and act in the world.

This matters because discourse analysis is not simply close reading or grammatical analysis of text. It examines how language use reveals and reproduces power relationships, social norms, assumptions, and ways of thinking that are often taken for granted by the speakers or writers themselves.

What Discourse Analysis Actually Studies

Discourse analysis can be applied to almost any form of communication: interview transcripts, media coverage, policy documents, everyday conversation, social media posts, institutional reports, or historical texts. What unites these applications is the underlying question: what is this language doing, beyond its literal surface meaning, and what does the way it is constructed reveal about the assumptions, power dynamics, or social context it comes from?

For example, a discourse analysis of news coverage of an economic policy might examine not just what the articles literally say, but which voices are quoted and which are excluded, what metaphors are used to describe economic actors, how passive versus active voice is used to assign or obscure responsibility, and what unstated assumptions the coverage takes for granted as common sense.

Major Approaches to Discourse Analysis

Critical Discourse Analysis (CDA) is explicitly concerned with power, ideology, and inequality. CDA researchers examine how language use in institutions, media, and politics reproduces or challenges existing power structures - for example, analysing how a government's language around immigration policy constructs particular groups as threats or as deserving of sympathy, often through specific, identifiable linguistic choices.

Conversation Analysis focuses closely on the structure of naturally occurring talk - turn-taking, pauses, interruptions, repair (how speakers correct misunderstandings), and other structural features of real conversation. This approach is less concerned with broad social power and more with the fine-grained mechanics of how interaction actually works moment to moment.

Discursive Psychology examines how people use language to construct versions of events, justify actions, and manage their own identity and accountability in interaction - often applied to interview data where researchers are interested in how participants actively construct meaning rather than simply reporting pre-existing facts or attitudes.

Foucauldian Discourse Analysis, drawing on the philosopher Michel Foucault, examines how discourse constructs what counts as truth, knowledge, and normal behaviour within a given historical or institutional context - often applied to studying how a concept like "mental illness" or "criminality" has been constructed and has shifted historically through the language used to describe and categorise it.

Choosing the right approach depends entirely on your specific research question - CDA suits questions about power and ideology, conversation analysis suits questions about interactional structure, discursive psychology suits questions about identity and accountability in talk, and Foucauldian approaches suit questions about how knowledge and categories have been historically constructed.

How to Actually Approach a Discourse Analysis Project

Start with a genuinely specific research question, not a broad one like "how is X discussed in the media." A workable discourse analysis question is narrower: "how does government press coverage frame the responsibility for youth unemployment," for example, is specific enough to guide a focused, manageable analysis.

Select your data corpus deliberately, not arbitrarily. Discourse analysis typically works with a relatively small, carefully selected set of texts examined in real depth, rather than a large sample analysed superficially. Be prepared to justify why you selected these specific texts and this specific time period, since this choice significantly shapes your findings.

Read your data multiple times before coding anything. Discourse analysis requires close, repeated engagement with the text to notice patterns in word choice, framing, and structure that are easy to miss on a single read-through.

Look for what is absent as well as what is present. A key technique in discourse analysis is noticing what is not said, which voices are not included, and what alternative framings were available but not used - these silences are often as analytically significant as what actually appears in the text.

Ground every claim in specific textual evidence. A common weakness in student discourse analysis is making broad interpretive claims without pointing to the specific linguistic features in the text that support them. Strong discourse analysis quotes and closely examines specific phrases, word choices, or structural features to justify each interpretive claim being made.

A Common Misconception Worth Addressing

Discourse analysis is sometimes mistaken for simply summarising the content or opinions expressed in a text - what people said. It is not that. Discourse analysis is interested in how language is constructed and what that construction reveals or accomplishes, which is a genuinely different and more analytically demanding task than summarising the surface content of what was said. A student summarising the opinions expressed in a set of interviews is doing content analysis, which is a different, legitimate method - but it is not discourse analysis, and reviewers familiar with the method will notice the difference quickly.

Discourse Analysis in Mixed-Methods Research

Discourse analysis is often assumed to be a standalone qualitative method, but it also appears regularly as one strand within a mixed-methods design - for example, pairing a quantitative survey measuring attitudes with a discourse analysis of how those same attitudes are expressed and justified in open-ended interview responses. In this kind of design, it is worth being explicit in your methodology about how the two strands relate: are they answering the same research question from different angles, or are they addressing genuinely separate sub-questions? A common weakness in mixed-methods dissertations is including a discourse analysis component without clearly explaining what distinct contribution it makes alongside the quantitative findings, leaving the two feeling like separate projects rather than one coherent study.

Getting Support With Your Discourse Analysis

Discourse analysis is a genuinely demanding method to apply well, since it requires both a clear theoretical framework and close, defensible engagement with your specific data. If you want feedback on whether your analysis is grounded closely enough in the actual text, or support choosing the right discourse analysis approach for your specific research question, CampusScribe's editors work with qualitative methodology chapters at every stage, from approach selection through to final write-up.

From choosing a topic to polishing a final draft, CampusScribe's subject-specialist editors and coaches help with dissertations, capstone projects, presentations, and more - whether you need topic guidance, structural feedback, or a final edit before your deadline.