Loading field map
Discourse analysis and conversation analysis are two closely related, often overlapping, traditions of research within communication studies and the social sciences more broadly. Both investigate language in use—not as an abstract system of grammar, but as a form of social action. Where they differ is in their primary focus, their methods, and their theoretical commitments. Discourse analysis typically examines how language constructs versions of reality, social identities, and power relations across stretches of text or talk. Conversation analysis, by contrast, focuses on the fine-grained, sequential organization of ordinary talk-in-interaction, asking how participants themselves produce and understand social order moment by moment. Understanding the subfield requires grasping both traditions, their shared roots, and the productive tensions between them.
At its core, this area of study asks a deceptively simple question: what do people do with language, and how do they do it? The answers matter because language is not merely a vehicle for conveying information. Through talk and text, people make requests, offer apologies, establish identities, assert authority, resist oppression, build solidarity, and accomplish countless other social acts. The stakes are therefore both empirical and political. Empirically, researchers want to describe the systematic methods by which social order is produced and maintained in everyday interaction. Politically, they want to show how language can naturalize particular ways of seeing the world, making contingent arrangements appear inevitable.
A central assumption shared by most work in this subfield is that language is fundamentally indexical: its meaning depends on context. The same words can perform very different actions depending on who says them, to whom, when, and in what setting. This insight leads to a methodological commitment to studying language in its natural context rather than through invented examples or controlled experiments. Another shared assumption is that social reality is not given but accomplished. Institutions, hierarchies, and even personal identities are not pre-existing facts that language merely reflects; they are ongoing achievements that language helps to create and sustain.
The modern study of discourse and conversation emerged in the mid-twentieth century from several converging sources. One important precursor was the linguistic tradition of pragmatics, particularly the work of philosophers like J. L. Austin and John Searle on speech act theory. Austin argued that utterances do not just describe the world but perform actions—promising, warning, naming, and so on. This idea that speaking is a form of doing provided a crucial foundation for later work.
A second precursor was the anthropological and sociological study of language in social life. The ethnography of communication, associated with Dell Hymes and John Gumperz, emphasized that speaking is a cultural practice governed by norms that vary across communities. This tradition drew attention to the diversity of communicative practices and the importance of studying language in its social and cultural setting.
The most direct intellectual ancestor of conversation analysis, however, was the sociological tradition of ethnomethodology, developed by Harold Garfinkel in the 1950s and 1960s. Garfinkel argued that social order is not imposed from above by rules or structures but is produced from within by members of society through their ordinary, practical reasoning. He studied the taken-for-granted methods people use to make sense of their social world. Conversation analysis emerged when Harvey Sacks, a student of Garfinkel, began applying these insights to the study of recorded conversation. Sacks, along with Emanuel Schegloff and Gail Jefferson, developed a rigorous method for transcribing and analyzing talk that remains the foundation of the field today.
Discourse analysis has a more diffuse lineage. It drew on British and Australian linguistics, particularly the work of M. A. K. Halliday on systemic functional linguistics, which emphasized how language choices are shaped by social context and in turn shape social meaning. It also drew on French poststructuralist thought, especially the work of Michel Foucault, who used the term "discourse" to refer to the broader systems of knowledge, power, and practice that produce what can be said and thought in a given era. These two influences—one linguistic, one philosophical—created a productive but sometimes uneasy alliance within discourse analysis.
Conversation analysis (CA) is the most methodologically unified tradition within the subfield. Its object of study is talk-in-interaction: ordinary conversation, but also institutional talk such as courtroom proceedings, medical consultations, and classroom interaction. CA's central claim is that interaction is orderly at a level of detail that participants themselves orient to, and that this order is produced through systematic practices that can be discovered through careful analysis of recorded data.
The method is distinctive. Researchers make audio or video recordings of naturally occurring interaction, then produce detailed transcripts that capture not just words but pauses, overlaps, intonation, and other features of delivery. Analysis proceeds by examining sequences of turns at talk, looking for recurring patterns and for evidence that participants themselves treat certain actions as relevant or problematic. A key concept is the adjacency pair: paired utterances such as question–answer, greeting–greeting, or invitation–acceptance/decline, in which the first part makes the second part conditionally relevant. Another key concept is turn-taking: the system by which participants allocate the right to speak, including rules about when transitions become relevant and how overlaps are managed.
CA has produced robust findings about the organization of repair (how participants deal with problems of speaking, hearing, or understanding), preference organization (how some responses are structured as preferred or dispreferred, such as acceptances versus refusals), and the design of turns to perform particular actions. In institutional settings, CA has shown how participants transform ordinary conversational practices to accomplish institutional tasks, such as a doctor managing a patient's concerns or a judge controlling a courtroom.
The strengths of CA lie in its empirical rigor and its insistence on grounding claims in the observable orientations of participants. Its limitations are equally clear. CA is deliberately agnostic about broader social structures such as class, gender, or race, focusing instead on the local production of order. Critics argue that this focus can obscure how power and inequality operate through interaction. CA practitioners respond that their approach can reveal how such structures are made relevant by participants themselves, but the tradition remains cautious about importing analyst-defined categories.
Discourse analysis (DA) is a more heterogeneous tradition, encompassing a range of approaches that share a concern with how language constructs social reality. Within communication studies, the most influential strand is often called critical discourse analysis (CDA), associated with scholars such as Norman Fairclough, Teun van Dijk, and Ruth Wodak. CDA examines how discourse produces and reproduces power relations, ideology, and social inequality. It typically analyzes texts—news reports, political speeches, policy documents, and the like—in relation to their social and historical context.
A central concept in DA is that language is not a neutral medium but a site of struggle over meaning. Discourses are ways of representing the world that make certain perspectives available and others unavailable. For example, a news report that describes an unemployed person as "lazy" and another that describes the same person as "a victim of economic restructuring" construct very different social realities and imply very different policy responses. DA seeks to make visible these constructive processes and to show how they serve particular interests.
Another influential strand is discursive psychology, developed by Jonathan Potter and Margaret Wetherell, which applies discourse-analytic methods to psychological topics. Discursive psychology treats psychological states such as attitudes, emotions, and memories not as internal mental entities but as discursive accomplishments—things people do in talk to manage accountability, justify actions, or construct versions of events. This approach has been particularly influential in the study of identity, where researchers examine how people position themselves and others through language.
DA's strengths include its critical edge and its ability to connect micro-level language use to macro-level social structures. Its limitations include a tendency toward interpretive flexibility: different analysts can produce very different readings of the same text, and the criteria for validating interpretations are less clear than in CA. Some critics also argue that DA can overstate the power of discourse, treating everything as text and neglecting material conditions or embodied experience.
The relationship between CA and DA is complex. They share a commitment to studying language in use and a rejection of purely formal or experimental approaches. Both traditions insist that language is social action. Yet they differ in fundamental ways. CA is primarily descriptive and empirical, seeking to uncover the methods of interaction. DA is often critical and interpretive, seeking to expose the ideological work of language. CA focuses on the sequential organization of talk; DA focuses on the broader discursive formations within which texts and talk are embedded.
These differences have produced genuine disagreements. CA practitioners sometimes criticize DA for being insufficiently rigorous, for imposing analyst categories onto data, and for making claims that participants themselves would not recognize. DA practitioners sometimes criticize CA for being naively empiricist, for ignoring power and history, and for treating the local organization of interaction as if it were the whole of social life. These debates are not merely academic; they reflect different visions of what the study of language in social life is for.
At the same time, there has been considerable cross-fertilization. Some researchers combine CA's attention to sequential detail with DA's concern for social structure, examining how institutional power is enacted through the micro-organization of talk. Others use CA findings as a resource for critical analysis, showing how apparently neutral interactional practices can have discriminatory effects. The boundaries between the traditions are porous, and many researchers draw on both.
The present field is characterized by methodological pluralism and increasing diversification. Digital communication has opened new frontiers, with researchers analyzing social media posts, text messages, and online forums using adapted versions of both CA and DA. Multimodal analysis has extended the focus beyond language to include gesture, gaze, and other embodied conduct, recognizing that interaction is not just talk. This development has been particularly important in CA, where video analysis has become standard.
Another significant development is the growth of applied work. Conversation-analytic findings have been used to improve communication in medical settings, to train professionals in interactional skills, and to design technologies that support human interaction. Critical discourse analysis has been used to examine political communication, media representation, and institutional practices, informing advocacy and policy debates. This applied turn has made the subfield more visible and more consequential, though it has also raised questions about the relationship between research and intervention.
The field also remains geographically and culturally diverse, though its institutional centers have historically been in North America and Europe. Researchers around the world have studied interaction and discourse in a wide range of languages and cultural settings, challenging universalist assumptions and enriching the empirical base. This diversity is not merely additive; it has forced the field to confront questions about the cultural specificity of its concepts and methods.
What endures across all this diversity is a fundamental commitment to taking language seriously as a form of social action. Whether the focus is on the precise timing of a pause in conversation or on the ideological work of a policy document, the underlying conviction is that language is not a transparent window onto the world but a constitutive part of social life. That conviction, and the methods developed to pursue it, define the subfield and give it continuing relevance.