Keywords

1 Introduction

Conversation Analysis (CA) is the theoretical and methodological framework that inspires the contributions to this volume. CA is an approach and methodology in the social sciences that is rooted in ethnomethodology (EM) and aims to describe, analyze, and understand interaction as “a basic and constitutive feature of human social life” (Sidnell 2010, p. 1). In addition to many other disciplines within the social sciences, researchers in applied linguistics, language teaching, and language learning have used this framework to make sense of the social organization of instructed-learning settings in the last three decades. Specifically, this edited volume uses ethnomethodological conversation analysis (EMCA) to: (1) develop a unified, emic (or participant-relevant) account of how members do classroom interaction in various contexts; and (2) explore how second language acquisition (SLA) research that uses CA methods (CA-SLA) can potentially be used to develop new, empirically grounded pedagogical implications by and for a broad range of language teaching professionals. In this context, we are well aware that this agenda is an enormously complex, not to say hazardous, undertaking (see Adams and Chen 1981, who note that 75% of innovations fail). Most importantly, we have to find a way or ways of reconciling the often widely divergent needs and interests of different segments of the community of language learning/teaching professionals who constitute our intended audience.

At the simplest level of analysis, our audience includes researchers, teachers, and those involved in teacher education. In this context, as Mori (this volume) notes, communication between researchers and teachers is often strained or non-existent; she attributes this situation to a long-standing dichotomy between theory and practice that has divided—and continues to divide—the applied linguistics and language teaching communities. While this preliminary characterization of the situation is perfectly correct as far as it goes, we would argue that the borders between theory and practice are in fact quite fuzzy, and that the communication problems that exist between different stakeholders are also more nuanced and complex than they might seem at first sight. So, to return to the question of who the stakeholders in our audience are: they minimally include researchers who are primarily engaged in the production of basic research with their own independent theoretical traditions, agendas and vocabularies, but also education specialists such as curriculum/materials designers, methodologists, testing/assessment specialists, and teacher educators and trainers. These specialists are also researchers in their own right, although the kind of research they do is likely more applied than that carried out by those engaged in basic research. So, we would argue that the first potential site for miscommunication or misunderstanding among stakeholders occurs when the different kinds of researchers mentioned above attempt to engage with each other. Finally, our audience also includes pre- and in-service teachers. While it is tempting to view teachers as practitioners who are merely recipients of different kinds of research, we should recognize that all language teaching professionals—including pre- and in-service teachers—profess and enact potentially quite different kinds of theories of language teaching/learning. Following Markee (1997a), teachers’ theories of language teaching and learning are often more experientially based and oriented to solving practical classroom problems than the more global or abstract issues typically addressed by researchers. However, teachers’ ideologies and belief systems are nonetheless profoundly theoretical in their own right and here too there is the likelihood that communication between researchers and teachers might not run smoothly.

What does CA have to say about the status of theory and these complicated matters of communication and applicability? As we will see shortly, applied CA can indeed provide some rather unique insights into these issues. However, we should first note that the application of CA methods to resolve practical problems in participants’ everyday lives is a comparatively recent development in the field. More specifically, it took about 20 years for CA writers to develop the confidence to look beyond the organization of ordinary conversation and systematically investigate the characteristics of institutional talk.Footnote 1 Furthermore, it took another 19 years before the next step in this evolutionary broadening of the CA agenda—namely, the idea that CA can and should be applied as an explicit tool of intervention and change in institutional talk—was taken by Antaki (2011). Finally, this trend is even more recent in CA-SLA work. While there is a now sizable CA-SLA literature on how classroom interaction is organized (see the sub-section on CA-SLA for details) only a few studies (see Rolin-Ianziti 2010; Sert 2015 and some of the contributions to the volume edited by Salaberry and Kunitz 2019a—see in particular Huth and Betz 2019; Kunitz and Yeh 2019; Lilja and Piirainen-Marsh 2019a, and Waring 2019) have so far explicitly addressed the potential implications of their own findings for teaching and testing (see below).

Second, we also argue that we may need to go beyond the parameters of current CA research outlined above if we are to understand how CA might successfully be used as an integrative tool of systematic change and intervention in second language (SL) curriculum/materials design, methodology and testing/assessment. More specifically, citing Antaki (2011) and Maynard (2006), Mori (this volume) argues that applied, interventionist CA work on institutional talk-in-interaction needs to take into account how larger contextual factors affect talk-in-interaction. In EMCA, this issue typically focuses on how broadly or narrowly the scope of context should be understood (see Kunitz and Markee 2016 for an overview). This is indeed an important issue, but here we extend the parameters of Mori’s discussion by arguing that CA researchers would do well to gain at least a consumer’s understanding of the diffusion of innovations literature, which is essentially concerned with how to make social change, including curricular innovation, actually happen (see Waring this volume; Pekarek Doehler this volume b; and Huth this volume; these authors engage with different aspects of this literature to varying degrees). More specifically, as we slowly develop an approach to pedagogy that is based on CA research, we must also take care to develop our understanding of how to package this innovation in ways that are likely to improve its chances of ultimate success.

In the pages that follow, we therefore review important theoretical and methodological concepts, and provide a brief overview of basic findings in the sociological EMCA and CA-SLA literatures. We then develop a similar overview of the diffusion of innovations literature as it pertains specifically to the management of CA-based curricular innovation.

2 Ethnomethodological Conversation Analysis

As mentioned above, CA is rooted in EM, which is a radical form of microsociology that is foundationally based on: (1) Garfinkel’s (1967) critique of Talcott Parson’s (1937) functional approach to sociology; and (2) his empirical respecification of philosophical ideas pioneered by phenomenologists such as Husserl (1960, 1970), and Schutz (1932/1967) (see also Majlesi this volume for further in-depth discussion of these matters). Essentially, EM aims to describe how participants in interaction make sense of each other’s actions. This goal is achieved with a range of different methods. For example, some exponents use broad ethnographic methods to understand the notion of context, while others (typically, EMCA researchers) rely on much narrower, cotextual (Halliday and Hasan 1976) or turn by turn analyses of talk. Furthermore, EM addresses a broad range of topics. These include work on cognition, institutional settings, and studies of work in the discovering sciences, to name a few.

On the other hand, the parent discipline of CA treats ordinary conversation as the “bedrock of social life” or, more technically, as the “primordial site of sociality” (Schegloff 1987a: 102). According to this perspective, ordinary conversation consists of the kind of mundane chit chat which friends and acquaintances engage in during the course of their everyday lives and is considered to be the baseline speech exchange system in talk-in-interaction (Sacks et al. 1974). The term “talk-in-interaction” was coined by Schegloff (1987b) and subsumes both ordinary conversation and institutional talk, which is viewed as a task- or context-specific adaptation of the practices of ordinary conversation (see Drew and Heritage 1992). Thus, CA research attempts to explicate the observable orderliness of all forms of talk-in-interaction by demonstrating how members orient to various practices as they engage in real time interaction. These practices specifically include turn taking, repair, sequence organization and preference organization, all of which are foundational to the social achievement of intersubjectivity.

In its endeavors, CA embraces the distinctive perspective of “ethnomethodological indifference” (see Garfinkel and Sacks 1970/1986; Psathas 1995) to a priori theory. As Markee (2008: 405) clarifies, this means that in CA, “emic theory emerges as a by-product of empirical analysis,” a position which reverses the theory-first, empirical analysis-second approach to knowledge construction that is normally adopted in etic (i.e. researcher-centric) research. This ethnomethological respecification of theory in emic terms obviously complexifies traditional etic boundaries between theory and practice in interesting ways. But more importantly, it again reminds us that the emergence of teachers’ experientially based theories of language teaching and learning is fundamentally located in their own observable teaching behaviors (see Sert this volume; Waring this volume).

Finally, a word about CA methodology is in order. Analysis is always based on audio, preferably video, recordings of naturally occurring interaction, which are then transcribed to highly granular standards. The transcription conventions first developed in the early 1970s by Gail Jefferson (see Jefferson 2004 for their most mature expression) focus on talk only and are widely accepted as the foundational point of departure for further analysis. More recently, Nevile (2015) has shown that the interest in embodiment in sociological CA has grown exponentially, as measured by the number of articles published in the journal Research on Language and Social Interaction (ROLSI) that now routinely include information about members’ eye gaze, gestures and other embodied behaviors in transcripts. At the same time, however, we note that no single set of embodied transcription conventions has yet achieved the wide-spread acceptance of Jeffersonian conventions in the field (though see Mondada 2016 and Goodwin 2018 as potential contenders). In the context of CA-SLA work, we observe that similar tendencies are at work. For example, almost all the contributors to this volume include some information about embodiment in their transcripts. Furthermore, the way they transcribe embodiment also varies greatly from chapter to chapter.

While CA researchers seek to push back the frontiers of knowledge, and therefore need highly granular transcripts as part of their methodological arsenal, the issue of the granularity of transcripts takes on novel significance in the context of models of CA-based teacher education (such as those posited by Sert this volume or Waring this volume), which may require teachers-in-training to transcribe their own classroom interactions. In such context, teachers-in-training are invited to use transcripts as practical, CA-inspired resources for their own professional development. Given the practical purposes of transcripts in teacher education (and on the basis of our own emerging, practical experience with it), we argue that, while it might be beneficial for teachers-in-training to be able to read transcripts in the original CA literature, it is unreasonable to expect teachers-in-training to produce Jeffersonian style transcripts of their own interactions with their students, particularly at the beginning of their studies. Consequently, we anticipate that designing a viable interface between the different forms of knowledge construction in which researchers and future teachers engage will prove to be a particularly delicate, ongoing task for teacher educators who wish to incorporate insights from CA into their teacher education programs. We return to these issues when we review the diffusion of innovations literature at the end of this chapter.

2.1 CA-SLA

CA-SLA is the offspring of two historically distinct fields of inquiry: sociological CA on the one hand, and SL studies/applied linguistics on the other. More specifically, CA-SLA researchers adopt the epistemology and analytical techniques of CA to study how participants empirically do language learning in real time (see Kasper and Wagner 2014 for a detailed overview of the range of issues that fall under the rubric of CA-SLA research). For present purposes, we concentrate on the subset of CA-SLA work which focuses on the kind of interactions that occur in second/foreign language classrooms.

In this context, we wish to emphasize that, of course, learning does not occur only in classrooms. Indeed, language-learning-as-use occurs just as frequently “in the wild” (see Hutchins 1995 for the origin of this construct). Good examples of this work on language learning behaviors (Markee 2008) in the wild may be found in Gardner and Wagner (2004) and, more recently, in the companion volume to the present collection (see Hellermann et al. 2019), which outlines the current state of the art regarding how language learning behavior is organized in the less overtly institutional context of the community. In short, these two perspectives on language learning are complementary rather than competitive. That is, while they certainly provide numerous insights into the different kinds of affordances for learning that occur in more or less institutional forms of talk-in-interaction, they also ultimately converge in blurring the lines between instructed and less guided forms of language-learning-as-use. This is evidenced by work which investigates how language problems first encountered in the wild may subsequently be brought back into the classroom for further pedagogical work (see Eskildsen and Theodórsdóttir 2017; Eskildsen and Wagner 2015; Lilja and Piirainen-Marsh 2019a, b; Wagner 2015).

As we have already noted, CA accounts of language learning behavior in any setting do not amount to a new theory of SLA. However, a powerful post-cognitive critique of a priori cognitive-interactionist SLA (see, for example, Long 1996) certainly does emerge as an important by-product of this body of work. Now, at first, CA’s lack of an a-priori learning theory that could predict and explain how learning processes function and why was regarded as a fundamental weakness by its cognitive-interactionist critics (see the special issues published in the Modern Language Journal in 1997, 2004 and 2007 to see how this kind of criticism evolved). However, by the mid 2000s, the elements of a CA rebuttal of this cognitive-interactionist counter-attack began to emerge. For example, Young and Miller (2004), Hellermann (2008), Pekarek Doehler (2010), Sahlström (2011) and Seedhouse (2010) (among others) started to reconceptualize the notion of language learning in social terms as a change in participation frameworks which becomes routinized over time. In this context, Markee’s (2008) paper on language behavior tracking (LBT) was instrumental in developing an emically longitudinal methodology for tracking how participants observably incorporate new language into their emerging interactional repertoires over time. More specifically, this LBT methodology involves two components: learner object tracking (LOT) and learning process tracking (LPT). More specifically, LOT is a technique which identifies language learning events that focus on specific objects of learning (or learnables; see Eskildsen and Majlesi 2018) during a particular time period. Meanwhile, LPT employs CA to describe how participants engage in a particular kind of language learning behavior.

This body of research has proved to be quite influential, as can be seen by the fact that most of the chapters in this book invoke this social, emically longitudinal perspective on language learning as a corner stone of the analyses that are presented in this volume. At the same time, the notion of longitudinality has itself undergone a good deal of complexification. More specifically, we can distinguish in the first instance between micro- and macro-longitudinal work; that is, between studies monitoring change (or lack thereof) over short versus long periods of time. Historically, micro-longitudinal work constitutes the older tradition (see, for example, Markee 1994, and for more recent studies, see Greer 2016; Kunitz and Skogmyr Marian 2017; and Sert 2017) but macro-longitudinal work has become equally important over the years (see, for example, Eskilden this volume; Hellermann 2008). Obviously, the distinction between these two approaches to doing CA-SLA is not absolute, as it is not always clear when micro-longitudinal work morphs into macro-longitudinal research. However, the distinction is a useful one because the methodological problems involved in maintaining a consistently emic perspective while doing macro-longitudinal work over months or years are probably more challenging than they are in micro-longitudinal work. Most obviously, in our experience, language learning related behavioral change that occurs over a few days is much easier to document emically than change that happens over a period of weeks, months or years. More specifically, demonstrating the extent to which participants themselves consistently and observably orient to such changes in their own behavior in the course of multiple, separate episodes of real time interaction over extended periods of time is, methodologically speaking, a very different matter from CA researchers being able to demonstrate such change from a post hoc, etic perspective. This issue is on the bleeding edge of CA-SLA methodology, and deserves much more attention in the future.

In addition, it is important to note that, in the field of EMCA, we may be witnessing the emergence of a critique of longitudinal studies (whether micro- or macro-longitudinally oriented) and of the conceptualizations of learning as positive (or positivistic) change that they embody. For example, Jakonen (2018) suggests that what he calls retrospective orientations to learning activities may offer new insights into the emic dynamics of language learning behavior. He argues that, if one can study how learners retrospectively refer to prior learning experiences (see also Can Daşkın and Hatipoğlu 2019a, b for what they call references to past learning events or RPLEs), it would then be possible to observe how “learners themselves ‘do learning’ by constructing change over time as opposed to being individuals to whom change merely happens” (Jakonen 2018, p. 4).

2.2 Teaching Interactional Competence

As we have already noted, it took a relatively long period of time for CA researchers to embrace the idea of applying CA findings in various contexts of institutional talk. This is all the more true for CA-SLA researchers: even those who have explored classrooms as language learning environments have typically been reluctant to discuss if and how their findings can have practical implications (and, most importantly, applications) for the teaching profession. Over the years, however, this situation has gradually changed, to the point that this entire volume is dedicated to studies that specifically engage in such discussion. The fact that we have asked all contributors to the volume to conclude their chapters with a section containing some, more direct, pedagogical implications of their own empirical work is the most obvious manifestation of this commitment. But let us backtrack a little and see how this budding interest in the pedagogical applicability of CA-SLA research came to life.

The first advocates of the importance of using CA findings for pedagogical purposes (see for example, Barraja-Rohan 1997) were concerned with the need to imbue language teaching, and more specifically the teaching of speaking skills, with empirical findings coming from research based on naturally occurring conversations in the L1. This position emerged as a reaction to model dialogues presented in textbooks (Wong 2002), which are typically produced at the service of vocabulary and grammar teaching but do not represent even close approximations of how participants in conversation interact with each other. So, the original idea consisted of adapting and applying what we know about interactional competence (IC) in the L1 to the teaching of IC in the L2. IC has been defined as the ability to produce recognizable social actions through timely and fitting contributions to the ongoing talk (see Pekarek Doehler this volume a; for an overview of different perspectives on the matter see the 2018 special issue of Classroom Discourse). In other words, IC has to do with participants’ ability to, for example, issue, accept or decline invitations and other similar social actions that are accomplished through the purposeful use of embodied language. In order to make ourselves understood by our co-interactants, it is not only the linguistic formulation of our turns that matters; their timeliness and their position in the unfolding talk are indeed equally crucial. This approach to speaking and, more specifically, to IC is clearly informed by CA’s view of language as action.

Initial attempts to translate research findings into instructional units for the teaching of L2 IC eventually led to the design of research-informed instructional materials that targeted either specific aspects of IC such as compliment sequences, request sequences, phone openings and closings, etc. (see for example Carroll 2011a, b; Huth and Taleghani Nikazm 2006; Olsher 2011a, b; Wong 2011a, b), or a combination of interactional features (see for example Barraja-Rohan 2011). With time, what looked like isolated proposals by individual researchers who happened to be involved in language teaching evolved into more systematic conceptualizations of the issues related to the implementation of CA-informed L2 IC instruction (Betz and Huth 2014; Salaberry and Kunitz 2019b). Eventually, the interest in CA-informed language teaching broadened to include CA-informed language testing (see for example: Kley 2019; Kunitz and Yeh 2019; Walters 2007, 2009, 2013). Furthermore, what started as a call to use CA findings to train teachers (Carroll 2010; Sert 2015, 2019; Wong and Waring 2010) has evolved into more encompassing enterprises that aim to provide CA-informed professional development for language instructors in charge of teaching L2 IC in their institution (see for example the innovative professional development initiative held at the Center for Languages and Intercultural Communication at Rice University in 2013–2018 under the direction of M. Rafael Salaberry). In fact, it has become increasingly clear over time that the idea of proposing CA-informed (or at least CA-inspired) language teaching and testing is not sustainable unless language teachers (and not just CA researchers who happen to be working within a language program) are also directly involved. Such engagement is not immune to a number of difficulties. Most importantly, these include: the amount of time that is needed to train teachers, complex methodological and practical issues related to making meaningful use of CA findings, and engaging with more institutional issues that have to do with who initiates and establishes the steps in the route of curricular innovation. It is thus to the literature on the diffusion of innovation that we now turn our attention.

3 Diffusion of Innovations Research

We begin this section by grounding the discussion that follows in our previous introductory comments regarding what counts as theory for different stakeholders in CA-SLA, applied linguistics, teacher training and language pedagogy. Drawing on Edelsky (1991), Markee (1997a) distinguishes between basic “THEORIES”/“RESEARCH” on the one hand and “theories”/“research” on SL acquisition/teaching on the other. The original example given in Markee (1997a) to illustrate what a “THEORIES/RESEARCH” perspective on SLA looks like was Krashen’s Monitor theory (MT). MT is clearly an example of an etic, quantitative approach to knowledge construction, which is produced in a top-down way by and for professional RESEARCHERS and whose pedagogical implications eventually trickle down to practitioners. In contrast, “theories” and “research” are typically emic constructs that are implemented by practitioner-researchers who use action research and other bottom-up approaches to knowledge construction. So, wherever we might (somewhat artificially, it seems) situate ourselves on the putative continuum of researchers and practitioners previously mentioned (see Mori this volume), the real problem faced by all stakeholders involved in the design, implementation and evaluation of CA-based pedagogy is how to use insights that are derived from both THEORY/RESEARCH and theory/research in a meaningful way.

As we will see shortly, achieving such a synthesis involves stakeholders actively engaging in packaging innovations in particular ways that promote rather than hinder their ultimate adoption. We believe that the literature on the diffusion of innovations not only provides potentially valuable insights into how to achieve such a goal, but also clarifies the complexity of the larger enterprise of making curricular innovation actually work. In this context, we also note that three contributors to this volume (Waring, Huth, and Pekarek Doehler b) have independently invoked important issues that are central to this diffusion literature.

In what follows we first show how Antaki’s (2011) work on interventionist CA provides a bridge to the literature on the diffusion of innovations. We then sketch out how one model of curricular innovation might be used as a consumer’s guide to understanding the kinds of issues that will likely have to be addressed in the on-going development of a CA-informed approach to pedagogy.

Antaki (2011: 9–14) notes that any attempt to engage in interventionist CA will likely run into the following problems: (1) whose perspective (for example, an institution’s or a client’s) is being advanced during an intervention? (2) what administrative power—if any—do outside CA consultants possess to make change actually happen? (3) what (often conflicting) interests and agendas do different stakeholders have, and how do they impact whether in the end change actually happens or not? (4) what do CA consultants need to know about the ethnographic context of an institution? and (5) why do potential adopters ultimately embrace or resist an innovation? In addition, Antaki (2011) briefly alludes to how difficult it is to plan and implement change, and correctly notes that CA consultants will have to take into account various moral, political and technical issues that may potentially have an impact on the ultimate success or failure of the innovations they design. Let us now see how these and other questions not discussed by Antaki may be subsumed and integrated into the model of curricular innovation developed by Markee (1997a, b; see also Filipi and Markee 2018).

Drawing on previous work pioneered by the language planner Robert Cooper (1982, 1989), Markee (1997a, b) proposes an ethnographic model of curricular innovation that is based on answering the following questions: Who Adopts What, Where, When, Why and How? Briefly, the Who section (which overlaps with Antaki’s Problem #3) deals with what roles different stakeholders play in the diffusion process. The range of stakeholders can be surprisingly large. For example, coming from an educational perspective, Fullan (1982) suggests that gatekeepers such as school superintendents, principals, deans, and heads of department may all be involved in determining whether an innovation is actually implemented or not, and also points out that parents and students may also play a crucial role in determining what happens to an innovation. Furthermore, following Lambright and Flynn (1980), these stakeholders tend to relate to each other as potential adopters (or resisters), implementers, clients, suppliers, and entrepreneurs (or, in our terminology, change agents). Note here the parallels between these categories and the ones used by Waring (this volume) in the diffusion of innovations-related coda to her chapter. Whatever categories we use, the important thing for change agents to remember here is that potentially large numbers of people may (sometimes unexpectedly) assert that they have a stake in deciding the fate of an innovation.

The Adopts section (which overlaps with Antaki’s Problem #5) focuses on the dynamic nature of potential adopters’ decision-making processes and highlights the fact they are frequently reversible. Thus, the take-away for change agents here is that, while potential adopters may initially have a favorable view of an innovation, they often change their minds in the longer term. Consequently, change agents must constantly be on the lookout for such vicissitudes in potential adopters’ decision making.

The What section defines curricular innovation as “… a managed process of development whose principal products are teaching (and/or testing) materials, methodological skills, and pedagogical values that are perceived as new by potential adopters” (Markee 1997b: 46, emphasis added). Note here that this particular definition adopts a language program director’s perspective on change, which may well be different from that of other stakeholders’ (in this regard, see also Antaki’s Problem #1). It also emphasizes that the newness of an innovation is a subjective matter of perception, not an objective fact. Finally, this section also subsumes the question of whether an innovation is initiated by insiders or outsiders (see also Antaki’s Problem #2). Briefly, inside and outside change agents have different rights and obligations: while inside change agents typically possess the administrative power to make innovations happen, outside change agents usually act as consultants who may advise clients on how to proceed but cannot enforce adoption.

The Where section (which overlaps with Antaki’s Problem #4) considers the context in which an innovation has to function. Following Kennedy (1988), sociocultural context is understood as an onion ring of cultural, political, administrative, educational and institutional variables that can potentially have an impact on classroom innovation, in which culture is held to be the most important variable. Note here that the scope of context as envisioned by Kennedy is broader than the kind of context that is discussed by Antaki. This insight is nicely illustrated by Huth’s (this volume) discussion of the gate-keeping role played by national and international organizations such as the American Council for the Teaching of Foreign Languages or the Council of Europe in developing new assessment standards. If the grand project of developing CA-informed assessment standards is to be ultimately successful, these organizations must be explicitly recognized as potentially important stakeholders in innovations that have a national or international scope.

The When section focuses on how long it takes an innovation to diffuse (note that Antaki 2011 does not address this question). The key issue to understand here is that innovation is not a linear process. More specifically, the rate of innovation typically starts out slowly, accelerates dramatically once a certain threshold of adoption is reached, and then slowly levels off. This observation is probably one of the most important practical insights for any would-be change agents attempting to develop and implement CA-informed pedagogy, as all innovations typically take much longer to diffuse than originally expected. Change agents must therefore learn to cultivate patience, while also looking out for opportunities to move the innovation process forward whenever they present themselves.

The Why section (which overlaps with Antaki’s Problem #5) is concerned with understanding the kinds of psychological profiles that different stakeholders have. People who are psychologically open to change have been shown to influence more conservative stakeholders, which is why innovation is not a linear process. Different kinds of people need different amounts of time to adopt (or perhaps ultimately reject) change. This section is also concerned with understanding a number of different properties that all innovations possess. The most important of these is the property of relative advantage (Rogers 2003), which has to do with whether potential adopters perceive an innovation to be beneficial to them or not (in this context, recall from our discussion of What-related issues that innovation is an inherently subjective process). Thus, to return to the idea that innovations need to be packaged in particular ways, it is important for change agents to actively understand why the innovations they propose may or may not be attractive to potential adopters. Such a project involves constant and effective communication among all stakeholders during the innovation process.

Finally, we come to the How section, which is concerned with understanding the advantages and disadvantages of top-down, bottom-up and hybrid approaches to change (note that Antaki 2011 does not address these issues). For present purposes, this section is probably the most important in the Who Adopts What … model in that it potentially provides us with important insights into how innovations may be packaged in different ways. Examples of these different models include the top-down Research, Development and Diffusion (RD&D) model; the bottom-up Problem Solving model; and the hybrid Linkage model (see Markee 1997b for details), which pragmatically synthesizes insights from the two previous models. The RESEARCHER-led RD&D model is widespread in academia and has a number of important advantages. It typically generates high quality innovations, which also tend to diffuse quickly, at least in the short term. However, it also suffers from some important disadvantages because implementers (i.e., teachers) are typically excluded from the development phase. As a result, they often lack ownership of such innovations. In contrast, the bottom-up, teacher-researcher-led Problem Solving model actively involves implementers in participating in the development of innovations from the very start of an innovation cycle. This characteristic promotes a high sense of ownership among researcher-implementers. However, the initial quality of innovations produced through the use of this model of change is often low, although this typically improves substantially over time. Finally, the Linkage model of change (hopefully) draws on the strengths of the previous two models and has the potential to achieve a synthesis of insights that are derived from both THEORY/RESEARCH and theory/research.

In this context, the SWEARFootnote 2 teacher education framework outlined by Waring (this volume) and Sert’s complementary proposals for IMDATFootnote 3 (see Sert 2015, 2019, this volume) are of particular interest in that they illustrate in practical ways how such a synthesis between CA THEORY/RESEARCH and theory/research might be achieved. More specifically, both models posit a constant back and forth between these different approaches to knowledge construction as the basis for ongoing, ethnographically-grounded reflection by teachers on how to improve the pedagogical utility of their own and their students’ interactional practices during classroom talk.

This being said, it is important to acknowledge that, in its (successful) attempts to gain a seat at the SLA table over time, most CA-SLA work to date has so far invoked a THEORY/RESEARCH approach to conceptualizing how IC is achieved. Such work is very valuable and obviously needs to continue, not least because not all CA-SLA RESEARCHERS are necessarily primarily interested in the pedagogical applicability of their findings. In this context, reflecting on her own particular intellectual background and experience, Pekarek Doehler (this volume b) effectively self-identifies as a RESEARCHER. At the same time, she also acknowledges that she has no particular expertise in pedagogy. She therefore correctly suggests that “we need a chain of experts so as to cover the many intricacies that pave the way between research into the nitty-gritty details of L2 development on the one hand and the enormous complexity of implementing measures for teaching or testing on the other” (Pekarek Doehler this volume b: 418, emphasis added).Footnote 4 We would like to conclude this discussion by noting that Pekarek Doehler’s call for collaboration between different kinds of experts has much in common with our own proposals for the necessity of a carefully packaged synthesis between THEORY/RESEARCH and theory/research as a prerequisite for successful CA-based curricular innovation. This insight underscores the interdisciplinary complexity that underlies on-going attempts to develop CA-based pedagogies. It also motivates our previous suggestion that stakeholders involved in the development of CA-based pedagogies would be well-served by gaining at least a consumer’s understanding of the diffusion of innovations literature as a useful resource for understanding innovation processes. However, we also recognize that embracing such a suggestion would entail stakeholders potentially engaging in a considerable investment of time and energy in getting up to speed on this literature, and many may well conclude that it is not in their best interest to do so. Thus, only time will tell whether (and if so, the extent to which) future RESEARCH/research on CA-based pedagogy will actually embrace this recommendation.

4 Conclusions

In this chapter, we have provided an overview of how CA-SLA has evolved from sociological CA. Our request to contributors to include a final pedagogical applications section in their chapters represents a pragmatic attempt to develop systematic links between THEORY and application. However, looking to potential developments in the future, we have also shown how the kind of applied CA predicated by Antaki (2011) potentially interfaces with important issues in the curricular innovation literature, and how at least three contributors to this volume (Huth this volume, Pekarek Doehler this volume b, and Waring this volume) have already begun explicitly to orient to such issues in their own work.

Finally, a brief word about how this volume is organized. It concentrates on four areas of CA-SLA. These include: (1) CA research in second language classrooms; (2) Research in Content-Based Language Classrooms; (3) CA Research and Teacher Education; and (4) CA and Assessment. In addition, two concluding chapters offer closing remarks on the four substantive sections mentioned above. In order to help readers navigate their way through this book, we also provide mini introductions to these topics at the beginning of each section so that readers can get a sense of how the various contributions hang together intellectually. We also anticipate that these mini introductions will help readers choose which chapters are of most interest to them and thus choose the order in which they read them. Lastly, we hope that this volume will contribute to the development of new directions in applied CA-SLA work and to the collaboration between different stakeholders in the attempt to develop cutting edge research and pedagogical practices.