Introduction

The aim of this chapter is to introduce teachers to the methodology of conversation analysis (CA); to explain what conversation analysis is, how CA research is undertaken and key concepts used, and to illustrate why a CA approach provides a useful resource to reflect on teaching practices in early childhood education. This introduction to CA provides a foundation for readers to engage with the chapters that follow, as all Handbook authors are using this approach to unpack and understand talk among children, or between children and early childhood professionals.

Conversation analysis is one of many approaches used to investigate teaching and learning interactions. Different methods and methodologies provide different lenses on how we talk with young children. The lens we choose determines what we see or find, creating particular resources to inform our practice (Reference DanbyDanby, 2002). The case we are making in this chapter – and through the evidence illustrated in the Handbook as a whole – is that research using the methods of CA allows us to see the mechanisms of interaction, and to see the interactional details of pedagogy. These details are particularly useful when applying recommended systematic approaches to teaching and learning such as notice, recognize, and respond (Reference Carr, Lee and JonesCarr, Lee, & Jones, 2004). Paying close attention to how we notice, recognize, and respond allows us to unpack teaching practices and offers insight into the practical achievement of concepts such as scaffolding (Reference Wood, Bruner and RossWood, Bruner, & Ross, 1976), guided participation (Reference RogoffRogoff, 2003) and sustained, shared thinking (Reference Siraj, Kingston and MelhuishSiraj, Kingston, & Melhuish, 2015).

Early childhood education is increasingly adopting a sociocultural approach, changing the ways we think about traditional ‘teaching and learning’ in more formal educational systems. Contemporary early childhood education encourages us to view infants, toddlers, and young children as competent and capable citizens who make a valuable contribution to their own learning. Curriculum frameworks have become increasingly popular in establishing learning outcomes for children from birth to adolescence. Frameworks offer guidance for learning in ways that take notice of the specific social and cultural contexts of children and teachers, often recommending a sociocultural approach to support children’s disposition for lifelong learning. The implementation of framework principles and practices, however, can be elusive. Using CA research to illustrate the everyday implementation of curriculum frameworks can fill in these details of how we engage in productive interactions with children. Essentially, the methods of CA provide accessible and transparent materials to reflect on the practices of learning and teaching in early childhood education and care (ECEC) (see Reference Bateman and ChurchBateman & Church, 2017a; Reference Church and BatemanChurch & Bateman, 2020).

What Is Conversation Analysis?

Conversation analysis is a wholly empirical methodological approach to understanding social actions, to provide ‘an inductive, micro-analytic, and predominantly qualitative method for studying human social interactions’ (Reference Hoey, Kendrick, de Groot and HagoortHoey & Kendrick, 2017, p. 151). CA is used to study how people make sense of one another, and how this sense-making is on display for participants (and analysts) in talk-in-interaction. ‘Talk-in-interaction’ is the term that conversation analysts often use to refer to the range of verbal, nonverbal, kinetic, and haptic resources people use to communicate and collaboratively achieve intersubjectivity (i.e.to arrive at some shared understanding). These terms, and other key analytic concepts used in CA research, are explained throughout this chapter.

The methodology and methods of CA emerged from the discipline of sociology at the University of California in the 1960s. It is here that Harvey Sacks began his development of a research methodology that became known as CA, and his legacy is one of being the founder of CA. He worked closely with colleagues Emmanuel Schegloff and Gail Jefferson to study how people interacted with each other and the mechanisms used to structure these interactions. Sacks, influenced by Harold Garfinkel’s work in ethnomethodology (the study of social interaction to see how people make sense of the world in orderly ways), started to pay very close attention to talk-in-interaction when researching telephone calls to a helpline. His primary interest was in how people organize social action rather than from a primary interest in language per se. In studying social actions, Sacks was able to show that deliberate and systematic features of ‘mundane’ everyday talk allowed people to do particular things and achieve particular actions (illustrated in the publication of his original lectures from the 1970s; Reference SacksSacks, 1995). Reference Sacks, Schegloff and JeffersonSacks, Schegloff, and Jefferson (1974) were able to specify the rules of interaction, showing how people take turns in everyday conversation: one turn at a time, one turn after another, and in a sequence of contingent turns. This ground-breaking paper was the first time that the systematic structure of conversation had been explained, identifying that speakers orient to these rules in order to make sense of one another.

Up until the 1960s, research in sociology had explored social practices as defined by particular categories (e.g. race, gender, age) and linguists mostly used hypothetical examples of talk to explain the usage of language. Sacks, Schegloff, and Jefferson were pioneers in demonstrating how social actions are achieved through language, and that these social actions have ordered, predictable, and rule-governed patterns of use. Harvey Sacks was interested in the talk of young children insofar as it provided insights into how people go about managing social actions, given that children are also participants in social interaction from birth. For example, his original work shows how children navigate restricted speaker rights (e.g., using the phrase ‘You know what?’ prompts an adult to respond ‘What?’, which then gives the child an opportunity to say something; see Reference SacksSacks, 1995, volume I, p. 256). He also wrote about how children construct stories (Reference Sacks, Gumperz and HymesSacks, 1972) and rules of games (Reference Sacks1995, volume I, p. 363), and the skills required in learning how to lie (Reference SacksSacks, 1995, volume 1, p. 565). More detailed discussions on the history and approach of CA are provided elsewhere in relation to ontology (Reference HeritageHeritage, 1984a); key principles (Reference Hoey, Kendrick, de Groot and HagoortHoey & Kendrick, 2017; Reference SidnellSidnell, 2010; Reference Sidnell and StiversSidnell & Stivers, 2012); practical guides (Reference Lester and O’ReillyLester & O’Reilly, 2019; Reference PsathasPsathas, 1995; ten Reference LeeHave, 2007); classroom interaction (Reference Edwards-Groves and DavidsonEdwards-Groves & Davidson, 2017; Reference Gardner, Sidnell and StiversGardner, 2012, Reference Gardner2019; Reference Mushin, Gardner and GourlayMushin, Gardner, & Gourlay, 2021); and early childhood education (for example, Reference BatemanBateman, 2015; Reference Burdelski and HowardBurdelski & Howard, 2020; Reference Mori, Zuengler, Martin-Jones, de Mejia and HornbergerButler, 2008; Reference Björk-Willén and CromdalChurch, 2009; Reference Danby, Baker, Hutchby and Moran-EllisDanby & Baker, 1998; Reference TheobaldTheobald, 2016).

Conversation analysis aligns with sociocultural learning theories that currently underpin and shape teaching practice throughout the world. From a CA perspective, the social and collaborative practices enacted by children and teachers are both the vehicle for and constitutive of learning. Rather than just focusing on either what the teacher or the child does, CA’s core concern is with the co-construction and choreography of social practices (see Reference GoodwinGoodwin, 2017). By paying close attention to how turns at talk are related to one another, this emergent approach to understanding the social world – now known as conversation analysis – has provided insights about communication that have deepened our understanding of human interaction. Speakers themselves use talk-in-interaction as a way of doing things and making sense of one another, so conversation analysts pay attention to the same things speakers pay attention to (pauses, gesture, word choice, intonation, hesitations, and so on) to see how human sociality is achieved (see Reference EnfieldEnfield, 2017).

How Is Conversation Analysis ‘Done’?

Conversation analysis aims to understand real interaction, usually captured through video or audio recordings. Rather than observing children in laboratory settings, observations are made in everyday settings such as early learning centres. Children and educators are not asked to respond to surveys, participate in interviews, or complete standardized tests, but rather to allow researchers to be present in the setting, to observe typical daily practices, and capture these practices on video. Recognizing the observer’s paradox (Reference LabovLabov, 1972) – the notion that the researcher’s presence can change the nature of the behaviour observed – researchers typically spend extended periods of time at their chosen research location, as ethnographers do, to mitigate the effect of introducing a researcher and their video camera to the early childhood setting. As CA is concerned with capturing interactional practices in situ, these video recordings serve as reliable documentation of the types of interaction that take place – at that time, with those participants, in that place.

Recording everyday interactions allows a study of these interactions through repeated reviewing of the recorded data to see how speakers make sense of one another (Reference Sacks, Atkinson and HeritageSacks, 1984a) and to see what speakers do, rather than what we think they do (Reference StokoeStokoe, 2013). Video recording is preferred to capture the embodied features of talk-in-interaction (see Reference MondadaMondada, 2018), especially when exploring interactions with infants and toddlers. Features such as gesture, eye gaze, and body position are regarded as just as important as talk in the co-production of social interaction. Rather than labelling embodied features of social interaction as generally ‘non-verbal’, CA is interested in how specific embodied gestures are intertwined with verbal features of interaction; video footage allows the researcher to re-visit how this is achieved in the observed interactions.

After the researcher has made a series of recordings in the same setting, CA requires very detailed transcription of the video observations (see Reference Jefferson and LernerJefferson, 2004). To this end, conversation analysts study the video recordings for sequences of interaction, to identify practices that may reoccur, providing a collection of a particular phenomenon (Reference Mori, Zuengler, Martin-Jones, de Mejia and HornbergerButler, 2008; Reference Björk-Willén and CromdalChurch, 2009), or alternatively a single case that illustrates how local social order is done (Reference BatemanBateman, 2021; Reference DanbyDanby, 2002; Reference SchegloffSchegloff, 1987). Approaching the task of transcribing video-recorded interactions with Sacks’ attitude of ‘unmotivated looking’ (Reference Hoey, Kendrick, de Groot and HagoortHoey & Kendrick, 2017; Reference SchegloffSchegloff, 1996), the goal is to identify what the participants themselves orient to in the interaction.

What Are Key Concepts in Conversation Analysis?

Once naturalistic (everyday) recordings are captured, and while transcribing these interactions in great detail, conversation analysts ask, ‘why this, in this way, right now?’ (Reference Seedhouse, Richards and SeedhouseSeedhouse, 2005, p. 251). The CA microanalytic transcription process offers opportunities to investigate how each turn builds on the prior turn of others to co-create sequences of interaction. As the researcher transcribes what was said and done by the participants in the video recording, the turn-taking between the participants becomes visible. By paying close attention to what each speaker does, we can see in detail how a conversation is co-produced through these turns of talk in everyday teacher-child interactions. For example, the following brief Extract 1.1 from a transcript of a recorded interaction between a four-year-old child (marked as CHD) and a teacher (marked as TCH) during a walk through a Welsh cemetery (Reference BatemanBateman & Waters, 2013: see Extract 3.1d in Bateman, Chapter 3, this volume) shows who spoke first (the child’s question), then who replied (the teacher’s explanation), and then the response to that reply (the child’s receipt of understanding).

The sequences of actions that are visible in turns at talk provide an ongoing display for speakers of how their talk is received by others (see Kidwell, this volume; Reference Kendrick, Brown, Dingemanse, Floyd, Gipper, Hayano, Hoey, Hoymann, Manrique, Rossi and LevinsonKendrick et al., 2020). This preoccupation with what speakers are doing in conversation (Reference Sacks, Atkinson and HeritageSacks, 1984b), is a helpful way to understand interaction as a series of actions.

The most basic sequence of actions is co-produced through adjacency pairs, where one turn (the first pair part) sets up an expectation for a particular type of next turn (a second pair part). For example, a first pair part shaped as a question should be followed with a second pair part in the shape of an answer. Likewise, a first pair part invitation should be responded to in a second pair part that accepts or declines the invitation. For example, a first pair part request for information (either by a child or a teacher) should be followed by a second pair part that provides the requested information, and so on. Reference Schegloff and SacksSchegloff and Sacks (1973, p. 298) explain that this basic sequence provides an in-built mechanism for monitoring understanding:

In CA research, we see that even very young children demonstrate an understanding of the interactional obligation set up by a first pair part, and may, for example, demand an answer from their mother if none is provided (Reference KeelKeel, 2016; see also Reference Kidwell, Sidnell and StiversKidwell, 2012, and this volume). The sequences of turns at talk provide researchers – and teachers – with immediately available data on what children are paying attention to, how they understand questions, instructions, or explanations by what they do next. It is this ‘nextness’ of conversation that displays speakers’ understanding of the prior talk; every next turn provides the vehicle for intersubjectivity (shared understanding). In other words, we can see how others understand each other from what they do next.

Where there is an alternative for types of second pair parts – for example, an invitation can be accepted or declined – one is typically preferred (done immediately, briefly, and to the point; e.g., accepting an invitation without delay) and the other is dispreferred (delayed and includes some sort of account as to why the preferred action is not done; e.g. declining an invitation after a pause, and explaining why the invitation cannot be accepted) (Reference Pomerantz, Atkinson and HeritagePomerantz, 1984). This action does not mean psychologically preferred (i.e. what the speaker wants) but rather is linguistically the most straightforward next action. So pervasive is this organizing principle of talk-in-interaction that people typically treat pauses of more than 0.3 seconds as a clue that a dispreferred action is coming next, because dispreferred actions are prefaced by some sort of delay. This preference organization is one of the mechanisms that contributes to the efficiency and progressivity of interaction (Reference Stivers and RobinsonStivers & Robinson, 2006). Children, parents, and teachers are sensitive to these organizing principles of talk-in-interaction, and the immediate context always determines the preferred next action.

Every turn is understood in relation to the surrounding turns at talk, and the ‘conditional relevance’ (Reference SchegloffSchegloff, 1968) determines what might be the most relevant next action at any given point. In other words, ‘the position of an utterance in a sequence is criterial to understanding what a turn at talk is doing’ (Reference CliftClift, 2016, p. 65). To help unpack this point, we return to the example above where the teacher’s turn in lines 2–4 (‘they’re not ↓a::rrows↓ >darlin< they’re called gra::ve- uh=↓gra::ve stones where they write the na:me of the↓per:son↓’) treats the child’s question as a request for information and provides a relevant next action explaining what the gravestones are for. We can see that the teacher’s response is the relevant next action – a fitting second pair part – by what the child does next: the receipt token ‘ah’ (see Reference Heritage, Atkinson and HeritageHeritage, 1984b) demonstrates a shift in the child’s understanding that the arrows on top (the gravestones) mark the name of the person who is buried there.

With each turn orienting to the prior turn(s), this in-built and highly efficient system is used to identify and resolve problems in talk-in-interaction. Defined as repair, speakers can at any point revise their own speech or query the meaning of another, either directly or, more commonly, by flagging some sort of trouble (e.g. ‘what?’, ‘huh?’; see Reference DrewDrew, 1997; Reference Kitzinger, Sidnell and StiversKitzinger, 2012). The rules of turn-taking enable confirmation checks at every point; where misunderstanding occurs, speakers can move immediately to repair the problem. Returning to our brief example (Extract 1.1), we see that the teacher manages repair at the beginning of the turn when explaining the function of gravestones:

The teacher self-repairs in her own turn (↓gra::ve- uh=↓gra::ve stones), the sort of self-revision that is common in conversation, where speakers fix or revise their talk within their own turn. Other-initiated repair seeks clarification of the prior turn by another person, usually providing an opportunity for the other speaker to fix whatever the trouble source might have been. In the example above, we see other-initiated repair, where the recipient (the teacher) identifies a repairable and goes ahead to fix it themselves (repairing the incorrect ‘arrows’ to correct ‘gravestones’). The way in which teachers do repair is important, because sensitivity is needed to support children’s continuing contributions to the ongoing interaction. CA’s interest in how repair is achieved can be very useful in unravelling the ways in which approaches to feedback can be tailored to the individual needs of each child.

Teachers are constantly shaping their talk to accommodate the level of understanding of the various children they interact with throughout the day, in what CA calls recipient design. In everyday life we try to avoid telling people what they already know, and shape the way we talk to meet the specific characteristics of the people with whom we are interacting. For example, early childhood teachers address parents in a particular way when they report news of the child’s daily accomplishments, which differs from how they interact with the children themselves, and speak differently again when they are socialising with friends. The ways in which we design our talk for specific recipients is key in teacher practice, where we are given the task of extending children’s knowledge whilst supporting their holistic wellbeing. Here the importance of knowing the child you are working with is imperative to building on their existing knowledge and extending their thinking in ways that are interesting for them. For this reason, talking with a group of children can be challenging, given the different knowledge and abilities of the recipients.

Difference in knowledge states can be described as the epistemic status of speakers (i.e. what each person knows; see Mushin & Gardner, Chapter 7, this volume). When the teacher asks the child a (recipient-designed) question, the child will demonstrate their knowledge (or lack thereof) in their second pair part response. Education research often identifies three-part sequences in classroom interactions which consist of initiation-response-evaluation (Reference MehanMehan, 1979), where the teacher usually produces a question (initiation), the child provides an answer, or best guess (response), and the teacher then evaluates the answer by providing some kind of feedback (evaluation). Within these sequences, then, is a display of what children understood the question to mean, and a display of their knowledge in the response they provide. In CA research we see how a teacher’s third turn can move beyond evaluation and expand on children’s responses in order to build learning trajectories (Reference LeeLee, 2007; Reference McHoulMcHoul, 1978; Reference WaringWaring, 2015). CA’s treatment of sequences of turn at talk enable us to understand the collaborative exchanges between children and teachers that support learning.

In essence, conversation provides a type of assessment-in-interaction, where children’s knowledge and sense-making are on display in each subsequent turn at talk. Paying close attention to how these turns at talk unfold provides teachers with insights into the collaborative work of talk-in-interaction. Importantly for early childhood teachers, the CA approach equips us with an awareness that every interaction is made up of sequences of actions. Our turns as teachers are significant when we consider how we might respond to children’s interests in ways that extend and support their learning. Through watching video footage of teacher-child interactions and transcribing them to clearly mark out the turn-taking evident within the sequence of action, we gain access to the orderly ways in which children and teachers achieve effective pedagogy in interaction. Essentially, we can see how teaching and learning is managed through turns at talk.

Why Is Conversation Analysis Useful for Early Childhood Education?

Through the CA process of collecting video footage of everyday interactions between children and early childhood teachers and engaging in the transcription process, paying attention to each turn at talk, we can see how it is that speakers make sense of one another by what is done in each next turn. This sense-making – made by the participants – is available to onlookers, be they analysts or teachers paying close attention to the interactional practices of pedagogy. This entirely data-driven, empirical focus of CA lends itself to research in early childhood education, because we can see what both the teachers and the children are doing to collaboratively construct teaching and learning interactions.

The fact that the original data and analysis is on display in CA research means that the method can be used as a vehicle for applied studies and professional learning. Current work by Elizabeth Stokoe and her colleagues at Loughborough University has ‘ ... demonstrated to great effect how useful the methods of CA are for communication ... ’ training for police, mediators, medical receptionists, and university admissions staff. Specifically, they developed the Conversation Analytic Role-play Method (CARM; see www.carmtraining.org), and have found that practitioners gain a deeper understanding of the interactional practices of their workplaces if they are supported to find the practices themselves, by exploring the sequential organization of talk (Reference StokoeStokoe, 2013, Reference Stokoe2014b; Reference Stokoe, Sikveland, Fors, O’Dell and PinkStokoe & Sikveland, 2017). Professional learning, in the form of reflection on practice that is directly informed by the details of practice, is relevant and constructive for current early childhood teachers (see Reference Church and BatemanChurch & Bateman, Reference Church and Bateman2019, Reference Church and Bateman2020). Using evidence from CA research enables teachers to re-visit the mechanisms of learning interactions that enable children to contribute their ideas, explore meaning, and extend concepts.

Conversation analysis’s approach to interaction as an entirely collaborative activity makes it a useful methodology for understanding the children’s role in the active co-construction of learning activities. Recognizing that all parties shape the context of the ongoing interaction means that CA is compatible with early childhood’s concern with children’s rights (Reference TheobaldTheobald, 2019) in that it does not privilege the contributions of particular speakers (i.e. limiting analysis to what the teacher does). A CA approach considers how all members of a group negotiate and co-construct opportunities to participate – including how exclusion might be done. The methods of CA emphasize the interdependence of talk, illustrating that each turn at talk is contingent on what other speakers do, or do not do.

Given that CA underscores the inherently collaborative and cooperative nature of talk-in-interaction, this research methodology implicitly pays attention to children’s agency and competence. Practitioners and researchers in early childhood education are necessarily concerned with listening to children’s voices and the essential contributions they make. Early childhood research seeks to involve children as active co-researchers (Reference ClarkClark, 2017; Reference Danby and FarrellDanby & Farrell, 2004; Reference Mason, Watson, Ben-Arieh, Casas, Frønes and KorbinMason & Watson, 2014; Reference Mukherji and AlbonMukherji & Albon, 2018), to understand children’s culture from a child’s point of view (Reference Clark and MossClark & Moss, 2001; Reference CorsaroCorsaro, 2017) and recognize children as agents in constructing the social context of early learning environments (Reference James and ProutJames & Prout, 1997). Because CA is interested in emic perspectives – that is, the participants’ own attention towards the ongoing activity – it aligns with research and educational policy that positions children as capable and agentive in interactions in early learning environments.

This emic orientation also allows us to detail children’s own practices, to see how it is that they order and interact in the local social context, and demonstrate their interests to others. Researchers in early childhood education are not members of the group they seek to understand; even experienced early childhood practitioners cannot claim to see the world from the perspective of a three-year-old child. CA research does not make assumptions about practice or categorize actions from the analyst’s point of view, but instead seeks to describe participants’ own methods for managing interactions with others. The video recordings, detailed transcription, and analysis enable phenomena relevant to children’s own lives to appear. As summarized by Kidwell, ‘the matter of what children “mean” or “intend” by their actions might seem to pose problems for researchers of children’s early communication. CA provides a unique and effective tool set for finding and evidencing their concerns, goals and motivations in interaction’ (Reference Kidwell, Sidnell and StiversKidwell, 2012, p. 518).

Conclusion

This chapter has outlined what CA is, how it is done, and its usefulness for early childhood education praxis. The methods of CA give an apparatus to explain the how of pedagogy in different contexts. The chapters in this Handbook can be described as ‘applied CA’, in that they explore practices in education settings, as a type of institutional practice (see Reference Antaki and AntakiAntaki, 2011; Reference Drew and HeritageDrew & Heritage, 1992; Reference Heritage and ClaymanHeritage & Clayman, 2010), and consider the implications of interactional practices for the work of teaching and learning. This type of insight is not possible through recollection alone, because we do not remember all the details of talk. It is close analysis that uncovers which of these details are important when talking with children – for example, pausing at opportune moments to allow children time to think and shape their response, designing questions in particular formats that prompt exploration of a concept, or extending topics in ways that are contingent on children’s own interest and knowledge.

Through understanding the systematic ways in which social interaction is achieved, CA can provide a magnifying tool to explore in detail the pedagogical practices that we engage in with infants, toddlers, and young children in our everyday interactions. CA allows us to hit the ‘pause’ button to explore in detail what children know and to reflect on how we might respond to their knowledge in ways that align with their interests. Through detailed transcriptions of child-teacher interactions, we see what children do in constructing social action, and also see their understanding of – and active contribution to – the social rules of educational settings, the peer-initiated and designed rules of play. We see children’s conceptual knowledge on display, visible through unfolding turns at talk. As social beings, we rarely consider just how it is that we manage to navigate multiple social interactions throughout our daily lives. We tend to underestimate how complex social worlds are. We move from one activity to the next with little consideration of how each turn at talk and gesture we perform adds to our own socialization and the socialization of others. CA transcription and analysis affords the luxury of such pause and reflection, as we seek to understand the interactions that build our social worlds.