Discourse Analysis: Language, Power, and Context

Discourse analysis treats language as social practice

Discourse analysis examines how language organizes meaning within a particular setting. It asks more than what a text says. The analyst may study how speakers name people, assign responsibility, present a claim as certain or uncertain, and make some responses appear natural while leaving others difficult to imagine. The method is useful for policy documents, interviews, news coverage and institutional records, but the question and material must be defined before patterns are interpreted.

A discourse is not a single word or a hidden intention. It involves recurring ways of describing a subject, the positions available to speakers and the practices in which those descriptions matter. A study of how an organization describes clients, for example, should attend to the organization’s procedures and audience as well as its choice of labels. Language can shape a decision without proving that every reader understood it in the same way.

Define a bounded corpus

State the question, the texts included and the reason for selecting them. Identify authors or speakers, intended audiences, genre, institution, period and any relevant events. A comparison of policy documents from two dates requires an explanation of what changed between them; a comparison of speeches and internal memos must account for their different purposes. Describe the search or sampling method so another reader can see which voices may be missing.

Keep a record of the corpus. Document versions, dates and the context in which each text circulated. If an interview is transcribed, note conventions that affect interpretation, such as omitted pauses or translated speech. An excerpt gains meaning from what came before and after it. Avoid relying on a striking quotation while ignoring a contrary passage in the same document.

Choose an approach appropriate to the question. Some analyses concentrate on close reading of categories, metaphors and assumptions; others compare patterns across a larger set of texts. A count of terms can support the inquiry but does not explain how a term functions. The same word may express endorsement, criticism or quotation depending on its context. Explain how passages were selected for detailed analysis and what would challenge the proposed reading.

Examine how a text works

Look at naming and categorization. Does a document describe people as “service users,” “applicants,” “burdens” or “risks”? What role does that description give them? Analyze agency in sentence structure: who is said to act, who receives an action and whose choice disappears in passive language. Modality such as “must,” “may” or “could” can signal authority and uncertainty. Metaphor and repeated contrasts can make a policy direction seem inevitable even when alternatives exist.

Attend to presuppositions: what must be taken for granted for a statement to make sense? A proposal to “restore compliance” assumes that compliance existed or that its absence is the central problem. Examine how the text supports that assumption and whether another framing would lead to a different response. Silence is harder to establish than wording. An omitted perspective matters only when the material and setting provide a reason to expect it.

Compare patterns rather than collecting isolated features. If a policy repeatedly assigns individual responsibility while describing structural barriers as background, show the pattern across several passages. Quote enough surrounding text to make the interpretation testable. Consider an alternative reading: a passive construction may conceal responsibility, or it may simply reflect the genre’s convention of foregrounding a procedure. Support claims about power with the institutional context, not with grammar alone.

Connect interpretation with consequences carefully

Imagine two policy documents about the same group. One calls members “service users” and describes routes to support; the other frames them chiefly as “risks” and emphasizes monitoring. An analysis can compare who appears to have agency, which problems are named and what interventions seem reasonable in each text. It should then check the documents’ purposes, dates and associated procedures. A change in language may accompany a change in practice, but the texts alone do not prove that services or outcomes changed.

Where a study claims real-world effects, seek other evidence. Implementation guidance, decisions, interviews or observations may show how staff and affected people used the categories. Distinguish a claim about the text’s available meanings from a claim about a speaker’s intention or a reader’s response. Those are different questions and require different kinds of support.

The analyst’s own position also matters. Explain why the subject and corpus were chosen, which concepts guide the reading and how contradictory examples were handled. Reflexivity does not require a personal confession; it makes the interpretive steps visible. If the researcher has access only to public documents, acknowledge the limits of inferring internal deliberation. If translated material is used, note choices that may affect terms central to the argument.

Write an analysis rather than a list of devices

Organize the discussion around a finding that answers the question. A paragraph can identify a pattern, present a short excerpt, explain the linguistic choice, relate it to the institutional setting and consider an alternative interpretation. Link subsequent paragraphs so the reader can see whether a pattern persists, changes or meets an exception. A table of terms can orient the reader, but the prose must explain why the contrast matters.

Avoid declaring that a text “constructs power” without specifying the mechanism and its limits. Power can concern authority to classify, ability to set an agenda or control over an institutional response. Name the relationship under study. State what the corpus supports and what would require evidence beyond it. This precision makes a critical interpretation stronger, not weaker.

Conclusion

Discourse analysis turns attention to the work language does in context. A credible study defines its corpus, examines patterns of naming and agency, connects them cautiously to social practice and tests its interpretation against alternatives. The result should show a reader how particular words and structures support a bounded claim, while leaving intention and real-world effects open where the evidence cannot establish them.