Interpretive Conversation Analysis
- Introduced by
- Harvey Sacks, Emanuel Schegloff, Gail Jefferson (CA foundations); interpretive extension by discourse scholars including Margaret Wetherell
- Year
- 1960s–1970s (CA); interpretive strand formalised 1990s–2000s
- Type
- Qualitative discourse research design
- Family
- Qualitative design / analysis
What does it do?
Interpretive conversation analysis (ICA) examines how meaning is co-constructed turn by turn in talk, combining the micro-sequential rigour of classic conversation analysis with an explicitly interpretive stance.
Rather than treating sequential organisation as the sole analytic object, ICA asks what participants are doing socially and discursively through their turns — what identities, institutional agendas, and power relations are built and contested in interaction. It draws on naturally occurring or recorded talk from social, institutional, or interview settings.
The intuition behind the method
Classic conversation analysis (CA) treats talk as a self-organising system and insists the analyst stick closely to what participants demonstrably orient to. Interpretive CA keeps that sequential discipline but opens the analytic lens: the researcher also asks what the interaction reveals about broader social meanings, institutional roles, or ideological work. Think of it as using CA's fine-grained toolkit — turn-taking, adjacency pairs, repair, preference — as a microscope, then stepping back to interpret what the organism is actually doing in its social environment. The analyst's own interpretive horizon is acknowledged and brought to bear, not suspended.
History
Conversation analysis was founded by Harvey Sacks, Emanuel Schegloff, and Gail Jefferson in the late 1960s, with Sacks's landmark lectures (1964–1972) establishing the turn-taking system and the sequential organisation of talk. The original tradition was strictly emic and resisted importing sociological categories. From the 1980s onward, scholars such as Paul Drew, John Heritage, and Anita Pomerantz extended CA into institutional settings, and discourse psychologists including Jonathan Potter and Margaret Wetherell developed discursive psychology, which brought an explicitly interpretive layer to interaction data. This interpretive strand — examining identity, ideology, and social action through sequential evidence — coalesced as a distinct orientation in the 1990s and 2000s.
When to use it
Use interpretive CA when you want to understand how meaning, identity, or institutional action is produced moment by moment in talk, and when a purely sequential CA account would leave the social significance under-theorised. It is well suited to research in health communication, educational interaction, organisational settings, and media discourse where both the fine grain of interaction and its broader social meaning matter. Appropriate data are recordings of actual talk — naturally occurring or elicited interviews. Do not use it when you lack recordings (transcribed summaries are insufficient), when the research question calls for attitude or belief measurement, or when statistical generalisation is required. ICA is not appropriate for purely structural CA questions that demand strict sequential analysis without interpretive overlay.
Limitations and considerations
- Requires high-quality audio or video recordings and careful transcription — fieldwork access and recording permission can be difficult to obtain.
- The combination of sequential and interpretive analysis demands substantial methodological literacy in both CA and discourse theory.
- Findings are context-specific and not statistically generalisable; transferability depends on analytic generalisability rather than sampling logic.
- The tension between CA's emic, data-driven ethic and the analyst's interpretive framework is never fully resolved and must be managed explicitly.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Grounds analysis in what participants themselves demonstrably do, providing a robust empirical base that purely interpretive approaches lack.
- Captures the sequential, real-time production of meaning in a way that post-hoc accounts (questionnaires, reflective interviews) cannot.
- Able to reveal how institutional power, identity, and ideology are enacted through ordinary interactional machinery.
- Flexible enough to combine with membership categorisation analysis, positioning theory, or critical discourse perspectives.
- Transcripts serve as transparent, auditable data that reviewers can examine to assess interpretive claims.
Applications
- Analysing how doctors and patients negotiate diagnostic authority in medical consultations.
- Examining how teachers construct student competence or failure through classroom interaction.
- Investigating how gender or racial identity categories are invoked or resisted in workplace talk.
- Studying how interviewers and respondents co-construct research data in qualitative interviews.
- Exploring how customer service agents manage institutional constraints in helpline conversations.
- Examining how political figures perform authenticity or deflect accountability in media interviews.
How it works
How it works
A set of recorded interactions is selected to address the research question — these may be naturally occurring conversations, institutional encounters (medical consultations, classroom exchanges, service calls), or research interviews. Sampling is purposive: episodes are chosen because they are rich sites for the phenomenon under investigation, not for statistical representativeness.
Audio or video recordings are transcribed using Jefferson notation or a simplified variant. Detail level is matched to the analytic focus: where sequential architecture matters, overlaps, pauses, and prosodic features are marked; where the goal is more broadly interpretive, a readable transcript with turn boundaries may suffice. Accurate transcription is treated as a first analytic step, not mere preparation.
The researcher works through the transcript to identify interactional sequences of interest — adjacency pairs, pre-sequences, repair initiations, topic transitions, and other sequential structures. Deviant cases (where the usual pattern is violated) are specifically sought as they often reveal the normative structure most clearly. This step preserves the CA commitment to participant-demonstrable orientations.
Beyond sequence, the analyst examines what participants are accomplishing: how institutional roles are enacted or resisted, how identity categories are invoked or avoided, how epistemic asymmetries are managed. Interpretive frameworks — such as membership categorisation analysis, positioning theory, or critical discourse concepts — are brought to bear to contextualise the sequential findings within broader social and institutional processes.
Findings from individual episodes are integrated into an analytic account of the interactional phenomenon. This account describes both the sequential machinery (how the talk is organised) and its interpretive significance (what participants are doing socially). The synthesis is illustrated with selected transcript excerpts that ground the interpretation in the data.
Common pitfalls
- Importing interpretive categories before the sequential analysis — treating identity or power as given rather than showing how they emerge in the talk risks overriding what participants themselves orient to.
- Thin transcription — omitting overlaps, pauses, or repair sequences sacrifices the sequential precision that distinguishes ICA from ordinary thematic analysis.
- Cherry-picking excerpts — selecting only episodes that confirm the interpretation without examining deviant or negative cases weakens analytic credibility.
- Conflating ICA with critical discourse analysis — CDA foregrounds macro-level power structures, while ICA derives its interpretive claims from micro-sequential evidence; merging them without clarity blurs both.
Frequently asked questions
How does interpretive CA differ from classic CA?
Classic CA insists the analyst stay strictly within what participants demonstrably orient to in the sequence, avoiding imported sociological concepts. Interpretive CA retains the sequential rigour but explicitly connects sequential patterns to broader social meanings — identity, power, institutional role — using interpretive frameworks. The difference is not in the sequential tools but in how far the analyst moves beyond the data to theorise social significance.
Do I need Jefferson notation?
If your analytic focus includes sequential features such as overlap, repair, or prosody, then Jefferson notation (or a clearly documented simplified version) is necessary. If your interpretive focus is primarily on meaning content and social action rather than sequential machinery, a readable transcript with turn boundaries may be sufficient — but you should justify this choice and acknowledge what is lost.
How many transcripts or episodes do I need?
There is no fixed number. CA-informed studies often analyse a small collection of episodes (5–30 instances of a phenomenon) in great depth. The criterion is analytic saturation — when additional episodes are not revealing new sequential or interpretive patterns. A single richly analysed case can be theoretically significant; breadth at the expense of sequential depth defeats the purpose.
Can interpretive CA be combined with ethnography?
Yes — this combination is common in studies of institutional interaction. Ethnographic context (fieldwork observations, institutional knowledge) can inform the interpretive layer of the analysis, helping the researcher understand what participants take for granted and what is being managed interactionally. The sequential analysis remains grounded in the recordings; ethnographic context enriches its interpretation.
Is interpretive CA appropriate for interview data?
It can be, but with caution. Research interviews are not naturally occurring talk — they are interactionally asymmetric and researcher-initiated. Applying ICA to interview data is legitimate when the research interest is in how participants construct accounts, navigate the interview as an institutional encounter, or co-produce narrative. It is less appropriate when interviews are used merely as a vehicle for accessing attitudes or experiences, where thematic analysis may be more suitable.
Related methods
Sources
ten Have, P. (2007). Doing Conversation Analysis: A Practical Guide (2nd ed.). Sage.
ISBN: 978-1412922271Copied.Wetherell, M., Taylor, S., & Yates, S. J. (Eds.) (2001). Discourse Theory and Practice: A Reader. Sage.
ISBN: 978-0761971566Copied.