top of page

Conversation Analysis 101

What is Conversation Analysis?

  • a research method for analyzing conversation (including both verbal and nonverbal aspects of social interaction)

  • focusing on spoken language, it seeks to uncover organization, structure, and rules of seemingly chaotic conversation

  • traces intersubjectivity: participants in a conversation co-create meaning together, with each turn made in the context of the previous turn and simultaneously creating context for the following turn

  • answers the question: "Why that, in that way, right now?" (Seedhouse, 2005)

  • introduced by Schlegloff, (1987) talk-in-interaction is the preferred term to conversation because it refers to two main domains of naturally occurring conversation CA distinguishes between:

1. ordinary conversation (i.e., between family members and friends)

(speech acts, syntax-for-conversation, reference, joke and storytelling)

2. institutional conversation (e.g., doctor-patient, teacher-student)

(news, medical, courtroom, classroom conversations)

  • collects data using audio and video recordings, which remain primary data sources researchers work with

  • transcribes audio and video recordings using elaborate systems of transcription symbols to capture as many details as possible because in CA no detail can be a priori ruled out as insignificant

  • it's emic, not etic approach requires a rich description of context, but the analysis is bottom-up and data-driven: a participant's rather than a researcher's perspective

  • the purist approach requires no prior theoretical assumptions, background or contextual details be considered unless participants themselves reveal orientation to it:

"A case is only convincing to the extent that it is directly motivated by the conversational data present for analysis" (Markee, 2000)

  • ethnomethodological, a subcategory of ACD (analysis of conversational data)

Some Key Principles of Conversation Analysis

1. CA is empirical

  • Data comes from how people use language

  • Attempts to explain data sequentially: Why does X follow Y?

  • Attempts to explain data according to distribution: Why does X occur with Y and not with Z?

2. Discourse includes linguistic form, meaning, and context

3. Form and meaning of everyday discourse is negotiated through interactions

4. Utterances are sequentially situated

5. How something is said, meant, and done is based on the following:

  • the speaker’s intentions

  • the sequential context of other utterances

  • the type of discourse (e.g. description, joke)

  • the social context

Some Conversation Analysis Terminology

TCU (turn-construction unit)............. the smallest amount of talk

TRP (transition relevance place)....... where speaker change may occur

turn-taking.......................................... who is allowed to speak and when

speech act .......................................... an utterance with a performative action

(e.g., promise, order, warning)

sequence............................................. a unit, such as an adjacency pair

adjacency pair ....................................a two-part speech exchange, where part 1 requires part 2 (e.g., in greetings Part 1: Hey! requires a response, Part 2: Hi!)

response token utterance that provides feedback (e.g. yeah, okay, uh-huh, oh)

assessment.......................................... an utterance that expresses hearer's sentiments

(e.g. good, that's fantastic, how terrible)

repair..................................................a correction made by the speaker

History of Conversation Analysis

  • Edmund Husserl and Alfred Schutz’s Phenomenological Approach

  • CA originated in the 1960s at the University of California in the works of Harvey Sacks, Emanuel Schegloff, and Gail Jefferson

  • Sacks studied rules and patterns of the sequence of actions, the structure of interaction, restricted to mundane, everyday conversation, and isolated from any other processes, free of impositions of any theories

  • CA developed on the basis of Erving Goffman's work on moral background of interaction and Harold Garfinkel's ethnomethodology, a method of sociological analysis on how people use everyday conversation to make sense of the world

  • later CA spread into psychology and linguistics

  • use of CA is a recent phenomenon (earliest examples in the field: Gaskill (1980) and Schwartz (1980))

Conversation Analysis and Second Language Education (SLE)

CA and Language learning and teaching (Seedhouse, 2005)

1. CA and applied linguistics, connection through action

2. CA and teaching language for specific purposes

  • comparison of teaching materials and reality, preparation of materials/task design

3. CA and teaching material design

  • comparison of authentic versus invented dialogue

  • direct application: explicit teaching of conversation phenomena

  • indirect application: materials include naturally occurring phenomena

4. CA and language proficiency assessment

  • can clarify advantages and disadvantages of assessment format and tasks

  • changes the definition of competence from static to dynamic, as co-constructed by participants in a given context

5. CA and NS versus NNS conversation

  • e.g. vowel marking, such as "uh" attached to word endings by Japanese speakers: pronunciation problem or interactional resource: "uh" as a discourse marker to buy time

Conversation Analysis in Second Language Acquisition (SLA)

CA and Language learning process (Seedhouse, 2005)

Three approaches:

1. ethnomethodological approach

  • neutral and agnostic in relation to learning theories and teaching methods in analysis

2. sociocultural theory approach

  • CA method is used as a tool in the service of other theories of learning CA for SLA

3. linguistic approach, a marginalized approach

  • quantitative treatment of CA findings

  • CA concepts and interactional constructs are decontextualized

  • ethnomethodological principles and the social action elements are absent


  • unlike many methods, CA includes transcripts of primary data

  • allows the reader to compare his/her own interpretation with that of the researcher

  • increases its validity by increasing the possibility of replication

CA and SLE

  • redefines language learners’ competence: although in the early stages many learners are linguistically limited, they are nevertheless interactionally competent

  • brings to focus the fact that both interaction and linguistic competence is co-constructed, not exclusively individual


  • time consuming because transcription of primary data is detailed

  • reliability of CA depends on: the selection of what is recorded, technical quality of recordings, and adequacy of transcripts

  • generalizability: CA works on the particular and the general simultaneously, as it seeks to abstract from the particular, the general, or the universal:

Is it possible to make significant, generalizable results? For quantitative researchers? For qualitative researchers?

  • if a distinction between ordinary conversations and other speech exchanges is made, then a more careful analysis of the differences among different speech exchange systems is necessary

CA and SLA

  • CA is a behavioral discipline, SLA is a cognitive one: whereas CA is suspicious of individual cognitive constructs, SLA seeks to describe cognitive processes that underlie the process of learning:

Can we use CA to analyze language acquisition processes that was originally made for analyzing language use? (Markee, 2000)

  • CA accounts for language use, not language acquisition

Sample Study

Barraja-Rohan, A. (2011). Using conversation analysis in the second language classroom to teach interactional competence. Language Teaching Research, 15(4), 479-507.

Objective To raise adult ESL students' awareness of the mechanism and the norms of spoken interaction

Participants Australia

and Setting Group 1: 20 adult Vietnamese immigrants/lower intermediate

Group 2: 10 international students/intermediate

  • enrolled in a 12-week English conversation course

  • both groups: little contact with Australian English L1 speakers

Procedure 1. Pre-instruction conversation

2. CA analysis of Pre-I conversation to determine a set of CA concepts to teach and inform the design of instruction during the course

3. instruction phase: listening/observation, cloze tasks, conversation manipulation, conversation reconstruction

4. Post-instruction conversation

5. CA analysis of Post-I conversation to determine if any improvement on given CA concepts had occurred

*Neither Pre-I nor Post-I conversation was a role-play, participants could choose both their partners and the topics

Findings 1. CA of Pre-instruction conversation revealed the need to teach:

  • response tokens to Group 1/Lower Intermediate Students

  • assessments to Group 2/Intermediate Students

  • adjacency pairs to both groups

2. CA of Post-instruction conversation revealed improvements in the use of the select CA concepts in both groups

Recommended Sources (in order of usefulness)

Articles and Chapters

1 Seedhouse, P. (2005). Conversation analysis and language learning. Language Teaching, 38(4), 165-187. doi:10.1017/S0261444805003010

Provides an excellent description of conversation analysis; particularly useful is its detailed survey of CA application in the fields of SLE and SLA.

2 Markee, N. (2000). Conversation analysis. Chapter 2. Mahwah, N.J.: L. Erlbaum Associates.

Chapter 2 provides a very concise description of conversation analysis and a very good discussion of applying CA in SLA.

3 Kasper, G. (2006). Beyond repair: Conversation analysis as an approach to SLA. AILA Review, 19(1), 83-83.

Provides a very concise definition and history of CA, as well as discussion of applying CA in SLA.

4 Gardner, R. (2008). 10 Conversation Analysis. The handbook of applied linguistics, 262.

Provides a concise discussion of CA's historical foundations, methodological issues, its principal findings, and some discussion of CA application and future directions.

5 Markee, N. (2007). Conversation analysis: Issues and problems. Boston, MA: Springer US

Provides a very good discussion of the weaknesses of conversation analysis.

6 Barraja-Rohan, A. (2011). Using conversation analysis in the second language classroom to teach interactional competence. Language Teaching Research, 15(4), 479-507.

Includes both a good introduction to CA as well as a sample study that illustrates one of the possible applications of CA in SLE.

7 Huth, T. (2011). Conversation analysis and language classroom discourse. Language and Linguistics Compass, 5(5), 297-309.

Provides a discussion of applying CA in SLE, focusing on language classroom discourse.

8 Goodwin, C., & Heritage, J. (1990). Conversation analysis. Annual Review of Anthropology, 19, 283-307.

A very good overview of CA from the anthropological perspective.


9 Have, P. t. (2007). Doing conversation analysis (2nd ed.). Los Angeles: SAGE.

A practical guide for learning and teaching conversation analysis as a method. Includes both theoretical basis and practical application of CA. Each chapter includes exercises and reading suggestions for further exploration of each topic.

10 Hutchby, I. (2008). In Wooffitt R. (Ed.), Conversation analysis (2nd ed. ed.). Cambridge: Polity.

Very useful for future practitioners; it includes many examples of CA.

Follow Us
Recent Posts
bottom of page