Discourse Analysis

Discourse analysis (DA) is a family of approaches for studying how language constructs meaning, identity, and power relations. It is not a single method with a fixed set of steps — it is a theoretically grounded practice that requires you to make deliberate choices about which tradition you are working in, what you are looking for in your texts, and why those textual features matter for your research question.

This page covers the three approaches most commonly used in the programs I supervise, with practical guidance on choosing between them, applying them, and writing up your analysis.

---

What Is Discourse Analysis?

At its core, discourse analysis examines how language does things — how it constructs social realities, establishes power relations, creates identities, and makes certain ways of thinking appear natural while rendering others invisible. All DA approaches share a basic premise: language is not a neutral window onto the world but an active force that shapes what we can think, say, and do.

Beyond that shared premise, however, the approaches diverge significantly. They differ on what counts as “discourse,” where to locate power, how systematically to analyze texts, and what kind of claims you can make. These are not minor technical differences — they reflect fundamentally different theoretical commitments rooted in different intellectual traditions.

The three approaches most relevant for thesis students in the social sciences and humanities are:

1. Critical Discourse Analysis (CDA) — focuses on how language reproduces or challenges power and inequality, with close attention to specific textual features
2. Discourse analysis in International Relations (IR DA) — examines how foreign policy discourse constructs identities, threats, and political possibilities
3. Poststructuralist discourse theory (PDT) — analyzes how broader systems of meaning are constructed, stabilized, and contested

---

The Main Approaches

Critical Discourse Analysis

CDA, associated primarily with Norman Fairclough and Ruth Wodak, analyzes the relationship between language use and social power. Its central claim is that discourse is both shaped by social structures and actively shapes them in return — a dialectical relationship (Fairclough (1992); Fairclough and Wodak (1997)).

CDA works at the level of specific texts. You examine concrete linguistic features — word choices, grammatical structures, metaphors, presuppositions — and connect them to broader social and political processes. Fairclough’s three-dimensional model is the most widely used framework:

1. Text — the linguistic features of the text itself (vocabulary, grammar, cohesion, text structure)
2. Discursive practice — how the text was produced, distributed, and consumed (genre, intertextuality, interdiscursivity)
3. Social practice — the wider social and political context that the text both reflects and helps to constitute

CDA is explicitly normative: it aims to reveal how language sustains inequality, and it treats this critical stance as a strength rather than a source of bias (Wodak and Meyer (2016)).

Best for: Research questions about how specific actors use language to legitimize, naturalize, or challenge power relations — for example, how policy documents frame migration, how political speeches construct national identity, or how media coverage reproduces racial or gender stereotypes.

Discourse Analysis in IR

Discourse analysis in International Relations draws on poststructuralist theory but has developed its own distinctive methods, particularly through the work of Lene Hansen and Jennifer Milliken. The central concern is how foreign policy discourse constructs the identities, threats, and relationships that make particular policies appear necessary or legitimate (Milliken, 1999; Hansen, 2006).

The key analytical moves in IR DA are:

• Identity construction — how “Self” and “Other” are defined in relational terms (a secure Europe defined against a threatening East, a civilized nation defined against barbaric enemies)
• Linking and differentiation — how identities are constructed through chains of association (linking democracy with peace, freedom, and progress) and opposition (differentiating civilization from barbarism)
• Policy as enabled by discourse — showing how certain policies become “natural” or “necessary” once particular identities and threats are discursively established

A key text is Hansen’s Security as Practice (2006), which provides both a theoretical framework and concrete research designs for studying how foreign policy discourse links identity and policy. The Copenhagen School’s securitization framework (Buzan, Waever, and de Wilde (1998)) is a related but distinct approach that examines how issues are discursively constructed as existential threats.

Best for: Research questions about how states, international organizations, or political actors construct identities, define threats, and legitimize foreign policy — for example, how EU enlargement discourse constructs European identity, how counterterrorism discourse defines the terrorist Other, or how humanitarian intervention is discursively justified.

Poststructuralist Discourse Theory

Poststructuralist discourse theory (PDT), rooted in the work of Ernesto Laclau and Chantal Mouffe (Laclau and Mouffe (1985)), takes the broadest view of discourse. This tradition draws on Michel Foucault’s foundational insights about the relationship between discourse, knowledge, and power, though Laclau and Mouffe depart from Foucault in important ways. For Laclau and Mouffe, discourse is not limited to language — it encompasses all social practices and institutions through which meaning is produced. The analytical task is to examine how systems of meaning are constructed, how they achieve temporary stability, and how they are contested.

Key concepts include:

• Nodal points — privileged signs around which a discourse is organized (e.g., “democracy,” “security,” “development”)
• Chains of equivalence — how different signs are linked together to form a coherent discourse (linking “freedom” with “market economy” and “individual rights”)
• Chains of difference — how discourses differentiate themselves from competing systems of meaning
• Floating signifiers — signs whose meaning is contested between competing discourses (e.g., “justice” means different things in neoliberal and socialist discourse)
• Hegemony — the process by which a particular discourse achieves dominance and its meanings come to appear natural

David Howarth’s Discourse (Howarth (2000)) provides an accessible introduction, while Jørgensen and Phillips (2002) offers a clear comparative overview of this and other approaches.

Best for: Research questions about how broader political projects or ideologies are constructed and contested — for example, how populist discourse restructures the meaning of “the people,” how competing discourses define “sustainability,” or how the meaning of “European identity” shifts across different political contexts.

---

Choosing an Approach

The approach you choose should follow from your research question, not the other way around. Each approach makes different assumptions and asks different kinds of questions. Use the decision points below to narrow your choice.

What is your unit of analysis?

• Specific texts and their linguistic features (speeches, policy documents, media articles analyzed at the level of words, sentences, and rhetorical structures) → lean toward CDA
• Foreign policy texts and the identities they construct (official statements, strategy documents, parliamentary debates analyzed for Self/Other constructions) → lean toward IR DA
• Broader systems of meaning across multiple sites (how a concept like “security” or “development” is defined differently across competing political projects) → lean toward PDT

What is your relationship to power?

• You want to critique how language sustains specific power relations or inequalities → CDA (which is explicitly critical and normative)
• You want to show how discourse enables certain policies by constructing identities and threats → IR DA (which is analytical rather than normative, though informed by poststructuralist critique)
• You want to map how competing political projects struggle over meaning → PDT (which focuses on hegemonic contestation rather than specific power abuses)

How fine-grained is your textual analysis?

• Very fine-grained — you will analyze specific word choices, grammatical constructions, metaphors, and rhetorical devices within individual texts → CDA
• Moderately fine-grained — you will analyze how texts construct identities and link them to policy positions, but without detailed linguistic analysis → IR DA
• Broader — you are mapping how signs are organized into systems of meaning across a range of texts and practices → PDT

---

What to Look For in Texts

One of the biggest challenges in discourse analysis is knowing what to look for once you are sitting in front of your texts. Each approach directs your attention to different textual features. The lists below are not exhaustive, but they give you a concrete starting point for your analysis.

CDA: Linguistic and Rhetorical Features

Drawing on Fairclough (1992) and Van Dijk (2008):

Feature	What to look for	Example
Vocabulary	Word choices that carry ideological weight; naming and labeling	“Illegal aliens” vs. “undocumented migrants” vs. “irregular arrivals”
Grammar and agency	Active vs. passive voice; who is the agent and who is acted upon; nominalization (turning processes into things)	“The police shot protesters” vs. “Protesters were shot” vs. “The shooting of protesters”
Metaphor	Figurative language that structures how we think about an issue	“Flood of refugees,” “war on terror,” “level playing field”
Presupposition	What the text takes for granted as already established truth	“We need to restore order” presupposes that order has been lost
Modality	Degrees of certainty, obligation, or permission (must, should, may, might)	“We must act now” vs. “We could consider acting”
Intertextuality	References to or incorporation of other texts and genres	A policy document quoting scientific reports to legitimize its claims
Cohesion and argumentation	How the text structures its reasoning; what evidence it invokes; how it builds toward conclusions	Causal claims, conditional reasoning, appeals to authority

IR DA: Identity and Policy Constructions

Drawing on Hansen (2006) and Milliken (1999):

Feature	What to look for	Example
Self/Other construction	How “we” are defined in opposition to “them”; what qualities are attributed to each	“The free world” vs. “rogue states”
Linking	How different positive qualities are chained together to construct an identity	Democracy + freedom + human rights + modernity = “the West”
Differentiation	How the Other is constructed through contrasting qualities	Civilized/barbaric, rational/irrational, peaceful/aggressive
Degrees of Otherness	Whether the Other is radical (fundamentally different) or less radical (different but recognizable)	Enemy vs. rival vs. competitor
Temporal constructions	How past, present, and future are narrated to justify policy	“We learned from Munich that appeasement fails”
Policy as natural consequence	How discourse makes specific policies appear as the only logical or moral response	“Given the threat, we have no choice but to intervene”
Spatial constructions	How geographic and political spaces are discursively constructed	“The European neighborhood,” “the arc of instability”

PDT: Discursive Structures

Drawing on Laclau and Mouffe (1985) and Jørgensen and Phillips (2002):

Feature	What to look for	Example
Nodal points	Central signs that organize the discourse; the key terms everything else connects to	“Sustainability” in environmental discourse; “security” in defense discourse
Chains of equivalence	How different signs are linked to produce a unified meaning	Freedom = democracy = market economy = the West
Chains of difference	How signs are distinguished from each other within a discourse	Differentiating types of migration: economic, political, humanitarian
Floating signifiers	Key terms whose meaning is contested between discourses	“Justice,” “development,” “the people”
Moments vs. elements	Signs whose meaning is (temporarily) fixed within the discourse (moments) vs. signs whose meaning remains open (elements)	In neoliberal discourse, “growth” is a moment; “equality” may be an element
Antagonistic boundaries	Where one discourse defines itself against another; the frontier between “us” and “them”	Populist discourse drawing a line between “the elite” and “the people”
Hegemonic articulations	Attempts to fix meaning and present a particular interpretation as universal or natural	“There is no alternative” (TINA)

---

When to Use It

Discourse analysis is appropriate when your research question is about how language and meaning shape social and political realities — not just what actors say, but how the way they say it constructs particular understandings of the world.

Use discourse analysis when:

• You want to understand how a particular issue, identity, or policy is constructed through language, not just described
• You are interested in what is taken for granted, left unsaid, or rendered invisible — not just what is explicitly stated
• You want to show how language enables or constrains political possibilities
• Your research question asks “how” something is discursively constructed rather than “why” something happened

Do not use discourse analysis when:

• You want to measure the frequency of themes or frames across a large corpus (that is content analysis or quantitative framing analysis)
• You need to establish causal relationships between variables
• You want to assess whether claims are factually true or false
• Your interest is in what people think or intend rather than what their language does (that points toward interview-based methods)

---

How to Apply It

While the specific analytical moves differ by approach, the overall process follows a common structure. The steps below apply to all three approaches, with notes where they diverge.

Step 1: Ground yourself in the theory

Before touching your data, read at least two foundational texts from your chosen approach. You need to understand the theoretical assumptions, not just the analytical techniques. Discourse analysis without theory is just impressionistic reading.

• CDA: Start with Jørgensen and Phillips (2002), chapters 3-4, then Fairclough (1992) or Wodak and Meyer (2016)
• IR DA: Start with Milliken (1999), then Hansen (2006), chapters 1-5
• PDT: Start with Jørgensen and Phillips (2002), chapter 2, then Howarth (2000)

Step 2: Build your corpus

Select your texts systematically. Your corpus should be:

• Relevant to your research question
• Bounded by clear selection criteria (time period, source type, topic)
• Justified — you must explain why these texts and not others

See Building a Corpus for detailed guidance.

Step 3: Develop your analytical framework

This is where you specify exactly what you will look for and why. Based on the “What to Look For” tables above, select the analytical categories most relevant to your research question. Your framework should:

• Name the approach you are using and its core assumptions
• Specify which textual features you will analyze (e.g., “I will analyze vocabulary, metaphor, and agency attribution following Fairclough’s three-dimensional model”)
• Explain why these features matter for your research question
• Define key analytical concepts in your own words, demonstrating that you understand them

Step 4: Read and code your texts

Read each text multiple times, each time with a different focus:

1. First reading — read for overall meaning and context. What is this text doing? Who produced it, for whom, and in what context?
2. Second reading — code for your selected analytical features. Mark vocabulary choices, metaphors, identity constructions, nodal points — whatever your framework specifies. Be systematic: go through each feature across each text.
3. Third reading — look for patterns across your coded features. What discursive patterns emerge? What is consistent? What shifts or contradicts?

Step 5: Identify patterns and construct your argument

Move from individual textual observations to broader analytical claims:

• CDA: Connect textual features to discursive practices (how the text was produced and circulated) and social practices (what power relations it sustains or challenges)
• IR DA: Show how identity constructions link to and enable specific policies; trace how Self/Other constructions remain stable or shift across texts
• PDT: Map the discursive structure — identify nodal points, chains of equivalence and difference, and antagonistic boundaries; show how meaning is fixed or contested

Step 6: Write up your analysis

Present your findings systematically, using direct quotations from your texts as evidence. Every analytical claim must be supported by textual evidence. See “Structuring Your Thesis” below for how to organize this.

---

Structuring Your Thesis

A discourse analysis thesis follows the same general structure as any thesis, but the analytical framework and findings chapters have specific requirements. Below is a chapter-by-chapter guide.

Introduction

Standard: research question, relevance, brief overview of approach and case, chapter outline. No special requirements for DA beyond making clear that you are studying how something is discursively constructed.

Literature Review

Situate your research question within existing scholarship. This includes both the substantive literature on your topic (e.g., EU enlargement, counterterrorism policy, climate change politics) and the discourse-analytical literature relevant to your area. Show what previous DA scholarship on your topic has found and where your thesis contributes something new.

Analytical Framework

This chapter is critical and must do three things clearly:

1. Specify your DA approach — State whether you are using CDA, IR DA, PDT, or a combination, and explain the core theoretical assumptions. Do not just name the approach — demonstrate that you understand its epistemological and ontological commitments.

2. Define your analytical categories — Specify exactly which textual features you will analyze and why. For example: “Following Fairclough’s three-dimensional model, I analyze vocabulary choices and metaphors at the textual level, intertextuality at the level of discursive practice, and the reproduction of neoliberal governance at the level of social practice.” Be explicit and precise.

3. Describe your corpus and method of analysis — What texts did you select, why, and how will you analyze them? Explain your coding process. If you have selection criteria, state them.

Findings / Analysis

This is the heart of your thesis. Present your analysis systematically, organized by theme, by text, or by analytical category — whichever structure best serves your argument. Key principles:

• Show your work. Use direct quotations from your texts as evidence. Do not just claim that “the discourse constructs migrants as a threat” — show exactly how it does so, with specific examples of vocabulary, metaphors, or identity constructions.
• Move between text and interpretation. For each observation, explain what the textual feature is, why it matters according to your framework, and what it tells us about your research question.
• Be systematic. Apply your framework consistently across all texts. If you analyze metaphors in one text, analyze them in all texts. Inconsistent application undermines your credibility.
• Connect micro to macro. Individual textual features are only interesting insofar as they reveal broader patterns. Always connect specific observations to larger discursive patterns and, ultimately, to your research question.

Conclusion

Summarize your findings, answer your research question, and discuss the broader implications. What does your analysis reveal about how discourse shapes the social or political phenomenon you studied? Discuss limitations honestly — including the limitations of your analytical approach and corpus selection.

---

Example from the Literature

To see what a well-executed discourse analysis looks like in practice, consider Lene Hansen’s Security as Practice (2006), one of the most influential applications of discourse analysis in International Relations. Hansen analyzes Western discourse on the Bosnian War (1992-1995) to show how representations of Balkan identity shaped — and constrained — the foreign policy options available to Western governments. By tracing how the conflict was discursively framed through competing constructions (the Balkans as “ancient ethnic hatreds,” as a “European” crisis demanding intervention, or as a site of genocide requiring moral action), Hansen demonstrates how each framing linked particular identities to particular policies. The “ancient hatreds” discourse, for instance, constructed Balkan peoples as fundamentally different from “civilized” Europeans, which made non-intervention appear reasonable — if the violence was rooted in centuries-old tribal enmities, outside intervention would be futile.

What makes Hansen’s study exemplary is its combination of theoretical rigor and analytical transparency. She specifies her analytical framework clearly (drawing on poststructuralist identity theory), selects her texts systematically (official speeches, parliamentary debates, and media commentary across multiple Western states), and shows her analytical work by grounding every claim in concrete textual evidence. The study illustrates how discourse analysis moves beyond simply reporting what actors said about Bosnia to revealing how the way they said it constituted identities and enabled or foreclosed specific policy responses. For thesis students, it serves as a model for how to structure a DA project from research design through to findings.

---

Common Pitfalls

Discourse analysis is one of the most frequently misapplied methods in student theses. Below are the most common problems and how to avoid them.

1. No theoretical framework. The thesis claims to do “discourse analysis” but does not engage with any DA tradition. The student reads texts and reports what they say, without any theoretical grounding for how they analyze the texts or why certain textual features matter. Choose a specific approach (CDA, IR DA, PDT), read its foundational texts, and build an explicit analytical framework that specifies your theoretical commitments and analytical categories. If you cannot name the approach you are using and explain its assumptions, you are not doing discourse analysis.

2. Confusing discourse analysis with close reading or summarizing. The findings chapter describes what texts say — their content and arguments — rather than analyzing how they construct meaning. The student summarizes rather than analyzes. Always ask “how does this text construct X?” rather than “what does this text say about X?” Focus on the mechanisms of meaning-making: the word choices, the metaphors, the identity constructions, the presuppositions. Your job is not to report content but to reveal how language works.

3. Being impressionistic rather than systematic. The analysis cherry-picks interesting quotes without a systematic process. The student finds striking examples that support their argument but does not apply their framework consistently across all texts. Apply your analytical framework to every text in your corpus using the same categories. Code systematically. Report patterns, including patterns that complicate or contradict your expectations. An analytical journal and consistent coding process will help.

4. Not connecting text-level findings to broader context. The analysis stays at the level of individual textual features without explaining what they mean for the broader social, political, or theoretical question. The student identifies metaphors or identity constructions but never explains so what. Every textual observation needs to be connected upward — to broader discursive patterns, and from those patterns to your research question and its social or political significance. CDA’s three-dimensional model is useful here even if you are not doing CDA: always move from text to discursive practice to social practice.

5. Treating discourse as a synonym for “what people say.” The student uses “discourse” to mean “discussion” or “debate” and never engages with the concept’s theoretical meaning. The thesis says things like “the discourse on climate change” when it simply means “what politicians say about climate change.” In DA, “discourse” has a specific theoretical meaning: a system of meaning that constitutes social reality. If you are using the term, use it precisely. If your analysis does not engage with how language constitutes reality (rather than merely reflecting it), you may not be doing discourse analysis — and that is fine, but then use a different method label.

6. Overambitious corpus. The student collects hundreds of texts and produces a superficial analysis of each rather than a deep analysis of a well-selected sample. Quality over quantity. A rigorous analysis of 15-30 well-chosen texts will produce a much stronger thesis than a surface-level reading of 200. Discuss corpus size and selection with your supervisor.

---

Key Readings

Introductory and Comparative

• Jørgensen, M. W., & Phillips, L. J. (2002). Discourse Analysis as Theory and Method. Sage. DOI: 10.4135/9781849208871 — Start here. The best single overview of CDA, PDT, and discursive psychology, with clear comparisons. Essential reading for any DA thesis.
• Gee, J. P. (2025). An Introduction to Discourse Analysis: Theory and Method (5th ed.). Routledge. — Accessible introduction with practical guidance on conducting analysis.

Critical Discourse Analysis

• Fairclough, N. (1992). Discourse and Social Change. Polity Press. — The foundational text for Fairclough’s three-dimensional model. Essential for CDA theses.
• Wodak, R., & Meyer, M. (Eds.). (2016). Methods of Critical Discourse Studies (3rd ed.). Sage. — Comprehensive overview of different CDA approaches and their methodological steps.
• Fairclough, N., & Wodak, R. (1997). Critical discourse analysis. In T. A. van Dijk (Ed.), Discourse as Social Interaction (Vol. 2, pp. 258-284). Sage. — Concise overview of CDA’s principles and methods.
• Van Dijk, T. A. (2008). Discourse and Context: A Sociocognitive Approach. Cambridge University Press. DOI: 10.1017/CBO9780511481499 — Van Dijk’s sociocognitive approach, which emphasizes the mental models mediating between discourse and society.

Discourse Analysis in International Relations

• Milliken, J. (1999). The study of discourse in international relations: A critique of research and methods. European Journal of International Relations, 5(2), 225-254. DOI: 10.1177/1354066199005002003 — Essential. Defines what DA means in IR and evaluates existing scholarship.
• Hansen, L. (2006). Security as Practice: Discourse, Identity, and the Question of European Defense. Routledge. DOI: 10.4324/9780203236338 — Essential. Provides both theory and concrete research designs for DA in IR. Chapters 1-5 are required reading for any IR DA thesis.
• Buzan, B., Waever, O., & de Wilde, J. (1998). Security: A New Framework for Analysis. Lynne Rienner. — Introduces the securitization framework; relevant if your topic involves how issues are framed as security threats.

Poststructuralist Discourse Theory

• Laclau, E., & Mouffe, C. (1985). Hegemony and Socialist Strategy: Towards a Radical Democratic Politics. Verso. — The foundational text for PDT. Theoretically dense; read Howarth (2000) or Jørgensen and Phillips (2002) first.
• Howarth, D. (2000). Discourse. Open University Press. — Clear, accessible introduction to discourse theory from a Laclau-Mouffe perspective. Good starting point for PDT.

Applied Examples

• Hajer, M. A. (1995). The Politics of Environmental Discourse: Ecological Modernization and the Policy Process. Oxford University Press. — Excellent example of discourse analysis applied to environmental politics. Shows how discourse coalitions form around shared storylines.

---

Related Methods

Discourse analysis overlaps with and can be combined with several other methods:

• Framing Analysis — Framing analysis examines how issues are presented and structured, often with a more systematic coding procedure than DA. If your question is more about what frames are used and how prevalent they are, framing analysis may be a better fit. If your question is about how language constitutes reality and power relations, discourse analysis goes deeper.

• Building a Corpus — Whatever DA approach you choose, you need a well-constructed corpus. This guide covers how to select, collect, organize, and manage your textual data for systematic analysis.