Textual analysis is a qualitative research method for systematically interpreting the meaning, structure and function of texts — written, visual or audio — to understand how they convey meaning and reflect the culture that produced them. Rather than simply summarising what a text says, you examine how it says it: the words, images, sounds, framing and silences a text uses to construct a particular version of reality. It sits at the heart of media studies, sociology, marketing, communication and literary research.
Use textual analysis when your research question is about meaning rather than frequency or cause — for example, how newspapers frame migration, how an advertising campaign constructs gender, or how a political speech builds authority. This guide explains what textual analysis is, the six main approaches and how they differ, a step-by-step process, a worked example, and the strengths and limitations every researcher should weigh.
What is textual analysis?
Textual analysis is an umbrella term for a family of qualitative methods that treat a text as the unit of study and ask what it means, how it produces that meaning, and what it reveals about the social world. In research terms, a “text” is anything that carries communicable meaning: novels, news articles, advertisements, films, photographs, social-media posts, song lyrics, interview transcripts, policy documents, packaging and even spaces such as museum exhibits. The method assumes that texts are never neutral. Every text is the product of choices — which words, which images, which order, what to leave out — and those choices encode values, ideologies and assumptions that the analyst’s job is to surface.
This is why textual analysis is fundamentally interpretive. The researcher does not stand outside the text counting features mechanically; they read closely, in context, and build an evidence-based argument about meaning. As the cultural-studies scholar Alan McKee puts it, textual analysis is a way of obtaining “an educated guess at some of the most likely interpretations that might be made of that text.” That educated, defensible, well-evidenced interpretation — not a single “correct” reading — is the deliverable.
“When we perform textual analysis on a text, we make an educated guess at some of the most likely interpretations that might be made of that text.” (Source: McKee, 2003, Textual Analysis: A Beginner’s Guide)
When and where is textual analysis used?
Textual analysis is used across the humanities and social sciences wherever meaning, representation or communication is the object of study. It is particularly common in:
- Media and communication studies — how news outlets frame events, how television represents social groups, how genres construct audiences.
- Sociology and cultural studies — how texts reproduce or challenge ideologies around class, gender, race and power.
- Marketing and consumer research — how advertisements, brand campaigns and packaging build identity, desire and persuasion.
- Literature and the arts — close reading of theme, narrative, imagery and form in novels, poetry, theatre and film.
- Politics and policy — how speeches, manifestos and official documents legitimise positions and construct “common sense”.
- Education and health — how textbooks, public-health campaigns or patient leaflets position learners and patients.
It is well suited to questions beginning with how and why: How does this campaign construct masculinity? Why does this coverage feel sympathetic to one side? It is less suited to questions about prevalence or generalisable cause, which call for quantitative designs. If you are still deciding between approaches, our guide to quantitative vs qualitative research can help you locate textual analysis within the wider landscape, and textual analysis is itself one of several methods of data collection and analysis you might combine in a mixed design.
Textual analysis vs related methods
Students often conflate textual analysis with neighbouring approaches, so it helps to draw the boundaries clearly. Content analysis is usually treated as one approach within textual analysis, but in its quantitative form it leans toward counting manifest features rather than interpreting deep meaning. Secondary research (such as a literature review) uses existing texts as sources of information, whereas textual analysis treats the text itself as the object of study — you analyse the document, not merely cite it. Discourse analysis can be seen either as a sub-type of textual analysis or as a distinct, more linguistically and politically focused tradition; the practical difference is how far you push into language, power and ideology. Keeping these distinctions straight in your methodology chapter signals to examiners that you have made a deliberate, defensible choice rather than reaching for a convenient label.
The main approaches to textual analysis (and how they differ)
“Textual analysis” is not a single technique but a toolkit. Choosing among its approaches depends on what you treat as the source of meaning — manifest content, latent themes, language-in-use, signs, persuasion or story structure — and on whether your stance is more descriptive or more critical. The table below compares the six approaches you are most likely to use in a dissertation.
| Approach | Core question | What it focuses on | Typical stance | Best for |
|---|---|---|---|---|
| Content analysis | What is present, and how often? | Manifest, codeable features (words, categories, images) counted systematically | Descriptive; can be quantitative or qualitative | Patterns across a large corpus (e.g. tone of 500 headlines) |
| Thematic analysis | What patterns of meaning recur? | Latent and semantic themes built up through coding | Interpretive, flexible | Identifying and interpreting themes across qualitative texts |
| Discourse analysis | How does language construct reality and power? | Language-in-use, framing, ideology, what is taken for granted | Critical, context-heavy | Power, identity and ideology in talk and text |
| Semiotic analysis | How do signs make meaning? | Signs, denotation/connotation, codes, myth (Saussure, Barthes) | Interpretive, structural | Images, advertising, visual and multimodal texts |
| Rhetorical analysis | How does the text persuade? | Ethos, pathos, logos, audience, argument and style (Aristotle) | Analytical, persuasion-focused | Speeches, opinion pieces, campaign messaging |
| Narrative analysis | How is the story structured and told? | Plot, characters, sequence, point of view, storytelling devices | Interpretive | Personal accounts, brand stories, news narratives |
The approaches overlap and are often combined. A study of a charity advertisement might use semiotic analysis to unpack the imagery, rhetorical analysis to explain its emotional appeal, and discourse analysis to situate it within wider narratives about poverty. Below is a closer look at each, with links to our dedicated guides where they exist.
Content analysis
Content analysis systematically codes the manifest (surface, observable) content of texts into categories, then examines patterns. It can be quantitative (counting how many of 400 news stories use the word “crisis”) or qualitative (interpreting the meaning of recurring categories). Its great strength is that it handles large corpora reliably and transparently. See our full guide to content analysis for coding frames and reliability checks.
Thematic analysis
Thematic analysis identifies, analyses and reports themes — patterns of meaning — across a dataset. Braun and Clarke’s (2006) six-phase model (familiarisation, generating codes, searching for themes, reviewing themes, defining themes, writing up) is the most widely cited framework. It is flexible across epistemologies and a natural fit when your texts are interview transcripts or open-ended survey responses. Our thematic analysis guide walks through each phase.
Discourse analysis
Discourse analysis treats language as social action. Rather than asking what a text says, it asks what the text does — how it positions people, naturalises assumptions and reproduces power. Critical discourse analysis (associated with Fairclough and van Dijk) is especially attentive to ideology and inequality. Use it when your question is about how language constructs identity, legitimacy or “common sense”. See our discourse analysis guide.
Semiotic analysis
Rooted in the work of Ferdinand de Saussure and Roland Barthes, semiotic analysis studies signs — anything that stands for something else. You distinguish denotation (the literal meaning of a sign) from connotation (its cultural associations), and trace how connotations harden into myth — the taken-for-granted meanings a culture treats as natural. It is the default choice for analysing images, advertising and other visual or multimodal texts.
Rhetorical analysis
Rhetorical analysis, descending from Aristotle, examines how a text persuades through ethos (credibility), pathos (emotion) and logos (logic and argument), in relation to a specific audience and context. It is ideal for speeches, editorials, manifestos and campaign messaging, where the persuasive design of the text is precisely the point.
Narrative analysis
Narrative analysis focuses on how stories are constructed and what work they do: the sequencing of events, characters, point of view, and the devices that make an account coherent or compelling. It is widely used with personal testimony, illness narratives, organisational stories and brand storytelling.
A step-by-step process for textual analysis
Although the approaches differ, most textual analysis follows the same five-stage workflow. Treat it as iterative — you will loop back as your reading deepens.
- Select your texts and define the sample. Specify the population of texts (e.g. all front-page articles on a topic from three newspapers over six months) and your sampling strategy. Purposive sampling is common: you choose texts that are information-rich for your question. State inclusion and exclusion criteria so the corpus is justifiable and replicable.
- Choose your approach and define the units of analysis. Decide whether you are doing content, thematic, discourse, semiotic, rhetorical or narrative analysis, and fix your unit — a word, sentence, paragraph, image, shot, whole article or whole campaign. The unit must match the question.
- Develop a coding frame and code the texts. Create categories or codes — either deductively (from theory) or inductively (emerging from the data) — and apply them consistently. Keep an audit trail of code definitions and example extracts. For quantitative content analysis, test inter-coder reliability; for interpretive work, keep memos that justify your readings.
- Interpret the patterns in context. Move from codes to meaning. What do the patterns reveal about how the texts construct their subject? Always read codes against the cultural, historical and production context — meaning lives in context, not in isolated words.
- Report with evidence. Present your findings as an argument supported by quoted extracts and described images. Show, don’t assert: every claim about meaning should be anchored to specific textual evidence the reader can inspect.
Worked example: analysing one short text end to end
To see textual analysis in action, take a single illustrative text — the opening of a tabloid news report:
“BENEFITS BRITAIN: Army of jobless cost taxpayers £5bn. Hard-working families foot the bill as claims soar to record high.”
1. Approach & units. We combine semiotic and discourse analysis. The units of analysis are the headline, the sub-clause, and individual loaded lexical choices (single words and noun phrases).
2. Units & codes (signs). Breaking the text into codeable signs:
- “Army of jobless” — a military metaphor; sign of mass, threat and invasion (code: them as horde).
- “cost taxpayers” / “foot the bill” — an economic-transaction frame (code: burden / drain).
- “Hard-working families” — a virtue label for the in-group (code: deserving us).
- “claims soar to record high” — upward-motion verb + superlative (code: crisis / spiralling).
3. Denotation vs connotation. Denotatively, the text reports a benefits-spending figure. Connotatively, “army” turns claimants into an advancing enemy, “foot the bill” casts them as parasitic on virtuous workers, and “soar” dramatises the number as an out-of-control emergency rather than a routine statistic.
4. Interpretation. Through a binary opposition — hard-working families versus an army of jobless — the text constructs welfare as a moral conflict between deserving taxpayers and an undeserving, threatening out-group. The discourse positions the reader inside the “us”, naturalising the assumption that benefit claims are a problem of individual failing rather than of economic structure. This is the Barthesian move from sign to myth: a contestable political claim is made to feel like plain common sense.
5. Evidencing the claim. Every step above is anchored to specific words in the text the reader can inspect — we are not asserting that audiences feel hostility, only showing how the text invites that reading. A reflexive note would add that an alternative outlet could frame the identical £5bn figure as evidence of a social safety net under strain, underlining that the meaning lives in the framing, not the number.
Strengths and limitations of textual analysis
Like every method, textual analysis is a set of trade-offs. Knowing them lets you defend your design in the viva and write a credible limitations section.
Strengths
- Depth of meaning. It reaches the how and why of communication that counting alone misses, surfacing ideology, framing and connotation.
- Unobtrusive and non-reactive. Texts already exist, so analysing them does not change participants’ behaviour as a survey or interview might.
- Rich, accessible data. Media, archives and online sources offer abundant, often free material, making it feasible for student budgets.
- Flexible and theory-friendly. It accommodates many theoretical lenses, from semiotics to critical theory, and scales from one text to a large corpus.
Limitations
- Subjectivity. Interpretation is inescapably shaped by the analyst. Two researchers may read the same advertisement differently, which raises questions about reliability.
- Limited generalisability. Findings illuminate the texts studied; they cannot be projected onto populations the way survey statistics can.
- It studies texts, not effects. Textual analysis tells you what a text offers, not how real audiences actually receive it — a different (reception) study is needed for that.
- Time-intensive. Close, careful reading of a sizeable corpus is slow and demanding.
The standard responses to the subjectivity critique are transparency and reflexivity. Transparency means documenting your corpus, coding frame and decision rules so others can follow your reasoning. Reflexivity means examining how your own position, assumptions and theoretical commitments shape your reading, and saying so explicitly. Together they convert “it’s just my opinion” into a disciplined, defensible interpretation — the hallmark of rigorous qualitative work.
Common mistakes (and how to do textual analysis well)
- Summarising instead of analysing. Describing what a text says is not analysis. Always push to how meaning is made and why it matters.
- Cherry-picking convenient extracts. Choose a systematic, justified sample and report disconfirming cases, not just the quotes that fit your argument.
- Ignoring context. A sign or phrase means different things in different cultural and historical settings. Anchor every reading in context.
- No coding frame. Without explicit, defined codes your analysis is unrepeatable. Define codes, keep examples, and (for content analysis) check reliability.
- Confusing the text with the audience. Don’t claim audiences “feel” something — claim the text invites or positions them to.
Do it well by stating your approach and units clearly, building a transparent coding frame, reading in context, evidencing every interpretive claim with specific extracts, and writing reflexively about your own role. If you are integrating textual analysis into a larger study, our research methodology guide shows how to position it within your overall design.
Struggling to design your textual analysis?
Our academics can help you build a defensible coding frame, analyse your texts and write a rigorous methodology chapter.
