Keywords

The term “classroom discourse,” as used in this review, encompasses language used by the teacher and the learners, teacher-learner and learner-learner interactions, as well as paralinguistic gestures, prosody, and silence. These linguistic and nonlinguistic elements are the observable dimensions of classroom discourse, and they constitute the bulk of earlier studies. Studies of classroom discourse have also explored sociocultural factors which play a critical role in shaping classroom discourse, including the participants’ socioeconomic and cultural backgrounds as well as their psychological and mental states. They constitute the unobservable dimensions of classroom discourse. Because of the limit of space, this review focuses on SL (second language) or FL (foreign language) classrooms and makes reference to L1 classroom discourse research only when it impacts SL/FL classroom research.

Early Developments

Research on classroom interaction and classroom events originated in the field of general education in the 1950s for teacher education purposes. It was motivated by the search for “objective” assessments of student-teachers’ performance in the classroom and the identification of “effective teaching.” The first major attempt was Flanders’ systematic analysis of classroom interaction (Flanders 1960). Influenced by Flander’s work, a plethora of SL/FL classroom interaction studies began in the 1960s, and a number of classroom discourse instruments based on Flanders’ system were drawn up for language teacher training purposes (see Allwright 1988). Early studies of SL/FL classroom interaction were also driven by the need to evaluate the effectiveness of the various FL teaching methodologies in the hope that the “best” method would be identified. The inconclusive findings, however, pointed to the problematic nature of the basic tenets of these studies. It was generally recognized that classroom processes were extremely complex and little understood. The aim of classroom-centered research, it was argued, should be descriptive rather than prescriptive. There was also a consensus that research should focus on both teachers’ and learners’ language and behavior.

Parallel to the development of research on SL/FL classroom discourse was the research on L1 (first language) classrooms. The impetus for research in this area came from the “language across the curriculum” movement in Britain in the late 1960s which drew attention to the important role of language in education. Inspired by the work of Vygotsky which emphasized role of spoken dialogue for children’s cognitive development (Vygotsky 1962), a number of studies has been conducted on L1 content classrooms (see for example, Barnes 1969), some focusing on specific aspects of the language used by teachers and learners, for example, the types of teacher questions and the learner responses elicited, the types of learner talk (“exploratory” versus “final draft”), and the mental processes reflected (Barnes 1969). Motivated by linguistic rather than educational concerns, Sinclair and Coulthard (1975) have proposed a grammar of spoken discourse, using the discourse of an entire lesson as their data set. Their descriptive framework, particularly their analysis of the hierarchical structure of discourse units, has been highly influential. Their proposed structure of the “exchange” as consisting of “initiating,” “responding,” and “follow-up” moves (IRF) has been widely adopted in both L1 and L2 classroom discourse studies.

Similarly, in the USA, the impact of teacher-student talk on student learning began to be widely recognized in the 1970s. Motivated by the conception of language as social action, classroom discourse and interaction were understood as the public enactments of social order co-constructed by discourse participants. The work of Mehan (1979) and his observation of the IRE (Initiation, Response, Evaluation) structure resonate with Sinclair and Coulthard’s IRF structure, though their points of departure are quite different.

Major Contributions

In reviewing the major contributions to classroom discourse studies, I shall group them under two headings: theoretical orientations and research approaches. This organization shows more clearly the paradigm shifts in both realms. Readers will find studies mentioned under one grouping often also appear under the other grouping, for obvious reasons.

Theoretical Orientations

Information Processing Theory

Until the mid-1990s, research on classroom discourse had been dominated by an information processing theory of learning based on an input–output model. Learning was understood as a process that took place inside the head of the individual, with little regard to its sociocultural contexts. The majority of studies focused on the analysis of language input, interactional processes, and language output, all of which were the observables in a classroom. Much of the research conducted in the 1970s and 1980s was “etic” (nonparticipant’s perspective) rather than “emic” (participant’s perspective) (for a review of research during this period, see Chaudron 1988). Typically, these studies were quantitative in orientation.

One strand of research on language input is teachers’ speech and how it affects learners’ language output. Early studies focused on the linguistic features of teachers’ modified speech to facilitate comprehension. Subsequent studies, however, have pointed out that interactional modifications resulting from the negotiation of comprehensible input are more important in facilitating language learning. Consequently, the research focus shifted to interactional structure and modification devices used by teachers to provide comprehensible input (Long 1983). The lack of evidence that comprehensible input produces higher quality learner output has led to the “Output Hypothesis” (see Swain 2005) which states that pushing learners to produce comprehensible as well as grammatically accurate output is important for language acquisition because it forces learners to process language at a deeper level and to notice the “holes” in their interlanguage. Subsequent research has further argued that an interactional process during which meaning is negotiated is particularly effective for language acquisition (Gass and Mackey 2006). The findings of studies on the relationship between negotiation of meaning and language acquisition have been somewhat inconclusive, however.

Another strand of research on language input is teachers’ questions and their corrective feedback. Adopting Barnes’ (1969) classification of teacher questions in L1 classrooms, mainly “open” versus “closed” and “pseudo” versus “genuine” questions, similar distinctions have been made between “display questions” (i.e., pseudo-questions) and “referential” questions (i.e., genuine questions). Referential questions have been found to elicit linguistically more complex responses from learners than display questions (Long and Sato 1983). The function of teachers’ feedback has been conceived as providing information for learners to confirm or disconfirm their hypotheses about the target language, and the notion of “error” has been reconceptualized from a developmental perspective (Allwright and Bailey 1991). More recent research has emphasized the importance of form-focused corrective feedback and “recasts,” or reformulations, though research findings have been inconclusive.

Earlier studies of learner output include learners’ turn-taking behavior and oral participation in different classroom settings. Learners who take more turns and hence generate more input, referred to as “high-input generators” (HIGs), were considered more effective learners than those who take fewer turns, referred to as “low-input generators” (LIGs) (Seliger 1983). This claim has been criticized for ignoring important factors such as the cultural backgrounds of the learners which could affect learners’ interactional behavior. Investigations of learners’ oral participation have examined the effects of learning arrangements and task types on learner participation. Pair and group interactions were found to generate more negotiation of meaning and a larger variety of speech acts than teacher-fronted settings. Tasks which required obligatory information exchange yielded more modified interactions and learner output in pair and group work than those where the information exchange was optional (Doughty and Pica 1986).

Sociocultural Theory

The shift in research paradigm in general education from information processing theory to sociocultural theory of learning, influenced by the work of the Soviet sociohistorical school (Vygotsky 1962), began in the late 1960s. However, it was not until the 1990s that this research paradigm began to make an impact on ESL research. This shift has led to a reconceptualization of language, context, and learning in profound ways. Sociocultural theory (SCT) of learning conceptualizes the relationship between the learner and the social world as dialectical rather than dichotomous and as mediated by cultural artifacts of which language is primary. Learners are not just passive recipients of language input and teachers are not just providers of input. Rather, the learners, the teacher, and the sociocultural context in which the discourse takes place are constitutive of what is being learned. Classroom discourse studies based on the input–output model have been criticized for presenting an impoverished and reductionist view of SL/FL learning.

A number of more recent classroom-centered studies have adopted key concepts in SCT as an interpretive framework for analyzing classroom discourse, including the Vygotskian concepts of zone of proximal development (ZDP), mediated learning and scaffolding. Classroom discourse has been reconceptualized as a major semiotic resource that mediates learning in the classroom (Lantolf and Thorne 2006; see also Thoms 2012 for a review of literature on socioculturally oriented studies on classroom discourse).

Teachers’ questions are no longer analyzed from the perspective of the type of response they elicit but from the perspective of how they mediate the co-construction of knowledge between teacher and learners. Responses to teachers’ questions are no longer just elicitations by the teacher but co-constructed by both the teacher and the learners. Learners are seen as participants in the co-construction of questions in IRF sequences. The interactions between the teacher and the learners are analyzed from the point of view of how both parties shape the way classroom tasks are defined and conducted. Adopting the notion of mediated learning, Swain (2005) extended the notion of “output” as external speech. She argued that external speech in collaborative dialogues is a powerful mediational tool for language learning because it encourages learners to reflect on “what is said” in language-related episodes while still being oriented to making meaning. Studies adopting the notion of scaffolded instruction in the learners’ ZDP noted that scaffolding can be mutual rather than unidirectional (i.e., from expert to novice) and can be provided by peers, even among very young FL learners (Lantolf and Thorne 2006).

The recognition of the importance of context in shaping the meanings of discourse for participants has also led to a shift from quantitative to qualitative analyses of classroom talk in content subjects. The understanding of classroom learning as “a journey through time for those involved” (Mercer and Dawes 2014, p. 436) has stimulated new research approaches such as “event history analysis” in which teachers’ and students’ questions are analyzed according to what comes before and after speakers’ “moves” and their impact on the unfolding discourse. Since then, there has been a growing interest in “dialogic teaching” (Alexander 2008), in which teachers encourage students to actively engage in exploring, extending, and deepening their own thinking and understanding. A number of studies have drawn on this notion to examine whether and how it opens up classroom talk and maximizes educational outcomes.

Research Approaches

Ethnographic Approach

The early classroom discourse research reviewed in the previous section mostly adopted a linguistic approach to data analysis with predetermined structural and functional categories. Typically, a priori analytic tool was chosen and applied to data analysis. Although in most cases, amendments would be made to the chosen tool in response to the data, the researcher already had preconceptions about the nature of the data and what categories would be relevant for analysis. The researcher, as an observer, had supremacy over the interpretations of data which consisted largely of the observable in the classroom. The voices of discourse participants were seldom heard. In the late 1980s and early 1990s, it became clear that studies of the observable in the classroom needed to be illuminated by the unobservable in the classroom. Researchers suggested that learners’ participation in the classroom could be affected by their learning styles, psychological states, cultural backgrounds, and beliefs about classroom behavior (Allwright and Bailey 1991). Studies of Asian learners’ participation in the classroom have noted that they are less willing to volunteer answers and take fewer turns than their non-Asian counterparts and that their observable behaviors are partly shaped by their cultural values and identities (Tsui 1996). Similarly, the way teachers pose questions and provide feedback and the kind of interaction they engage with learners are shaped by their conceptions of teaching and learning and their lived experiences of classroom events.

Subsequently, classroom research has adopted an ethnographic approach and has analyzed classroom discourse data in its sociocultural contexts from an emic perspective. This approach is typified by the researcher “spend(ing) an extended period of time in the community under investigation, participating either overtly or covertly in people’s lives, observing, listening, and asking questions in the data collection process to gain insights into the issues being studied.” (Tsui 2012, p. 383). The emic perspective in ethnographic approach is achieved through interviewing the participants in order to gain access to their mental and psychological states (See Tsui 2012 for a review of ethnographic approaches to classroom discourse studies.)

An ethnographic approach to classroom discourse analysis is motivated by an interest in the co-construction of educational processes by both the teacher and the students, how discourse processes open up or close down opportunities for learning, and what is being learnt. Studies have been conducted on the socialization of ESL learners into different learning environments and the difficulties and opportunities that they have encountered. The unraveling of the co-construction of knowledge and opportunities for learning by both the teacher and students has led to a shift from a static and deterministic view of the classroom as being shaped by context to an understanding of classroom and context as dialectically related. As such, the agency of the teacher and the students in co-creating opportunities for learning in the classroom is very important.

Ethnographic studies of classroom discourse typically focus on specific slices of classroom life and examine in detail locally managed interaction. As such, it has been referred to as “micro-ethnography” (see Garcez 2008 for a summary of such studies) and has close affinity with ethnomethodology or conversational analysis (see next section). More recent studies of classroom ethnography have a combination of micro- and macro-ethnography in which the wider educational and sociopolitical contexts have been taken into consideration to illuminate the classroom data (Duff 2009; see the subsequent section on critical classroom discourse analysis).

Conversational Analysis Approach

In recent years, a growing number of socioculturally oriented studies have adopted a conversational analysis (CA) approach, on the ground that a linguistic approach is not adequate in uncovering the complexities and the multilayered contingent interpretive acts of the emergent discourse in the classroom. Conversational analysis, originating from the work of ethnomethodologists (Schegloff and Sacks 1973), studies the social organization of talk-in-interaction in natural settings from the perspective of how talk is oriented to and accomplished by the participants in a specific context. In this sense, CA is emic in orientation: The analysis is done not from the perspective of the researcher but from the perspective of how participants understand and manage each other’s talk as displayed in their own talk. Fine grained analysis is conducted on the verbal and nonverbal interactions in conversations, not as individual acts but as social actions which are oriented to and managed by the participants. Aspects of context such as gender, race, and power will be included only if they are demonstrated to be relevant to the participants. Studies adopting a CA approach have focused on specific aspects of classroom discourse and have revealed “subtle interactional practices which transform our perceptions of L2 learners and teachers” (Seedhouse 2012, p. 1). For example, Hellermann (2007), adopting a conversational analysis approach, conducted a longitudinal study of the dyadic interactions of six successful learners in opening teacher-assigned tasks in ESL classrooms and examined how over time the same student dyads had incorporated the teacher’s and their peers’ language to manage the openings. Hellerman argued that the detailed conversational analytic approach to the longitudinal data has enabled us to understand language development as a change in the use of resources to accomplish a particular social action and has uncovered “how learners manage and adapt to the affordance (van Lier 2000) of this conversational practice as a site for second language development.” (p. 91) (see Seedhouse 2012 for a review of classroom discourse adopting CA approach).

Critical Discourse Analysis Approach

Critical classroom discourse analysis, coined by Kumaravadivelu (1999), was proposed in response to the limited and limiting insights provided by a (socio)linguistic approach to classroom discourse which sees discourse as merely contextualized language use in the self-contained mini-society of the classroom, with little attention paid to its broader sociocultural, sociopolitical, and sociohistorical dimensions. Drawing on insights from poststructural and postcolonial theories, Kumaravadivelu proposed adopting a critical perspective in which classroom discourse analysis should take into consideration power relationships among the discourse participants as well as their competing beliefs, values, identities, and voices. A critical perspective of classroom discourse, according to Kumaravadivelu, is transformative in that it enables classroom practitioners to reflect on and to respond to sociocultural and sociopolitical structures that impact directly or indirectly on classroom discourse.

Many of the more recently published classroom discourse studies have adopted a critical analytic approach to uncover the forces shaping classroom discourse through investigating issues relating to ideology, power, knowledge, class, race, gender, social positioning, and identities (see Coates 2012 and Blackledge 2012 for reviews of literature).

Critical classroom discourse analysis is especially prevalent in studies of multilingual classrooms where issues of power, identity, culture, and values are particularly palpable. The multilingual classroom is therefore an immensely rich site for the investigation of the processes of social and cultural (re)production and the relationship between micro classroom and macro institutional processes. In the 1980s, studies on multilingual classroom discourse provided insights on how multiple language resources were drawn upon by teachers and learners as they negotiated the daily classroom routines in complex communicative processes. Since the 1990s, an increasing number of studies have adopted a critical ethnographic approach and have tried to relate the micro-analysis of classroom discourse to the sociopolitical and ideological processes at language policy levels by drawing on social theory, poststructural, and postcolonial theories. For example, a number of researchers have studied code-switching in multilingual classrooms as a resistance to or as a way of managing the challenge of medium-of-instruction policies which construct the supremacy of English as the only legitimate language that can be used in the classroom (see Martin-Jones 2015 for a review of work in this area).

In addition to gathering classroom discourse data, studies in this strand typically gather data from policy documents and in social spaces beyond the classroom, for example, staff rooms, meetings, playgrounds, and also other institutional settings where language policy and curriculum issues are discussed. Martin-Jones (2015) pointed out that critical ethnographic research on language policy and critical classroom discourse research have become increasingly intertwined and theoretical insights have been drawn from both strands.

Work in Progress: Crossing Disciplinary Boundaries

Recent research on classroom discourse has tried to advance the field by adopting conceptual frameworks in a variety of disciplines, most of which are sociocultural in orientation.

Activity Theory

Classroom discourse studies have begun to draw on insights from activity theory (Engeström 1991) which conceptualizes goal-oriented human action as part of a larger activity that is driven by motive and also shaped by the broader sociocultural system in which the activity is situated in the classroom. The individual’s participation in these socially meaningful activities is mediated by the cultural tools which he or she appropriates. Classroom discourse is one of the cultural tools alongside material objects, audio and visual artifacts, which mediate the achievement of the goals of classroom activities. In the course of the interaction, the cultural tools, the nature of the activity, and the modes of participation are transformed; the same activity may be realized by different actions mediated by different tools. Conversely, the same action may be driven by different motives, hence realizing different activities. According to this perspective of learning, the same task may be operationalized as different activities with different goal-oriented actions by different learners and by the same learner in different contexts. The dialogic interaction that emerges in task completion plays an important part in shaping the way learners orient themselves to the task and to each other. It is the orientation of learners as agency, their values and beliefs, personal and collective experiences, and the way they connect to the current experience, not the task per se, that determines the way the task will be performed and the learning that will take place (Coughlan and Duff 1994). Hence, tasks should be understood as emergent interactions and not as the packaging of language input. Though there are still not many studies adopting activity theory as their theoretical framework, the framework is particularly powerful in relating micro classroom processes to the macro institutional or societal processes and the dialectical relationship between the two.

Ecological Theory

Also working within the sociocultural paradigm, some studies have adopted an ecological theory of language learning. For example, van Lier (2000) emphasized the totality of the relationships between the learner and all other elements or participants of the context with which he or she interacts. He proposed “affordance” as an alternative conception of “input” and pointed out that the environment makes available opportunities for learners to engage in meaning-making activities with others (a “semiotic budget”), and what is perceived as relevant and acted on by the learner becomes an “affordance” (p. 252). In other words, “input” has been reconceptualized as the linguistic affordances perceived and used by the learner for linguistic action. Input is therefore not something external to the learner waiting to be acquired but rather the interaction between the learner and the environment.

Social Theory of Learning: Learning as Social Participation

Classroom discourse research has also begun to draw on the conceptual framework of learning as social participation (Wenger 1998), which is sociocultural in orientation. For example, Donato (2004) distinguished “interaction” in the second language acquisition literature from the notion of “collaboration” in a social theory of learning which entails mutual engagement in a joint enterprise that is socially meaningful to members of a community of practice. He noted that the relational dimension of collaboration has been largely ignored in SL/FL classroom research. He argued that the analysis of discourse generated by isolated task completion in short time frames by group members who are new to group work does not capture the reality of how learning is co-constructed in collaborative work because it takes time to establish relationships. Drawing on the notions of “community of practice” and “legitimate peripheral participation,” Donato further maintained that the value of collaboration is not to enable learners to acquire more language knowledge but rather to move from peripheral to full participation as competent members in their communities of practice. Classroom discourse studies have also drawn on the concept of learning as changing participation in the community of practice to which the learners belong to elucidate the language learning process.

Language Socialization

Language socialization is concerned with how novices are socialized to use language and how they become culturally competent members in the target culture through language use in social activities. For example, Duff (1995), drawing on language socialization theory, studied the discourse in an English-immersion history classroom in Hungary and investigated the socialization of students from a transmissive mode to an open enquiry mode of learning and the learning opportunities that were opened up consequently. More recent research has conceptualized the classroom as a place where the expert and the novice negotiate not only knowledge and skills but also values, identities, positionality, epistemic, and affective stance. For example, Morita (2000) studied the socialization into academic discourse of the nonnative and native-speaker of English graduate students through their engagement in oral presentations. The findings suggest that the socialization process is not a unidirectional process of the enculturation of the novice into the community of experts but rather a complex process of negotiation of their identities and knowledge not only as novices in oral academic presentation but also as ESL professionals with expert knowledge. The dynamic and fluid process of negotiation is a recurring theme in classroom discourse socialization research. For example, He (2015) examined the discourse in Chinese heritage language classrooms and showed how in the process of explaining the semantic and structural components of the Chinese characters, Chinese culture, moral, and values were inculcated by the teacher and how in the socialization process, the learners articulated their identities and their own understanding of the culture values and meanings. He further showed that all participants in multilingual classrooms could be agents in the co-construction of identities and speech communities. Finally, recent research has also drawn attention to discourses outside the classroom that are relevant to classroom learning as they serve to socialize learners into discourse practices in the classroom as well.

Phenomenography and Variation Theory

Phenomenography is an empirical educational research approach which investigates how various aspects of, and the phenomena in, the world are experienced in qualitatively different ways by people. Marton and Booth (1997) argued that learning involves a change in the way a phenomenon (or an object of learning) is experienced, conceptualized, perceived, and understood. According to them, “learning proceeds, as a rule, from an undifferentiated and poorly integrated understanding of the whole to an increased differentiation and integration of the whole and its parts.” (p. viii). Hence, research on learning is focused on “the variation in ways of experiencing phenomena” (Marton and Booth 1997, p. 111), and on the “architecture of this variation in terms of the different aspects of that define the phenomena” (p. 117).

Phenomenographic work adopts mainly qualitative (and sometimes a mix of quantitative and qualitative) methodology. Much of this work analyzed data from learners’ account of their experiences and conceptions. However, more recent work analyzed discourse data in both L1 and L2 classrooms from the perspective of how variation in the learning experience is brought about by the discourse and the artifacts used by the teacher in the classroom. For example, Marton and Tsui (2004), through analyzing data from ESL, mathematics, history, economics, and Chinese language classrooms, showed how critical aspects of the object of learning can be varied and how learners’ awareness of the critical aspects of object of learning can be brought to the fore by the questions used by the teacher, the sequencing of questions and discourse moves, and the co-construction of exchanges between the teacher and the students, and among the students.

The number of studies adopting a phenomenographic framework is still small. However, the conceptualization of learning as seeing and experiencing things in a different way, and the theory of variation that underpins how learning is brought about, would provide the theoretical framework for examining proposals such as “focus on form” and “noticing” in second language acquisition as effective pedagogical strategies for acquiring grammatical accuracy.

Problems and Difficulties

From the above review of major contributions to the field, including work in progress, it is apparent that classroom discourse research has made significant progress in addressing issues germane to understanding the complex interplay between factors which impinge on what appear to be simple classroom interchanges. Research on classroom discourse in the last two decades has begun to move away from being “data-heavy but theory-light” (Donato 2004, p. 299). As we have seen in this review, the appropriation of research methods and theoretical frameworks in other disciplines has enriched our understanding of classroom discourse. However, the field is faced with a number of challenges of which only a few obvious ones will be outlined here. One challenge is whether there is a propensity to adopt methodologies without understanding their origins and theory-method relationships and to use the same terminology with different theoretical assumptions in the discussions. For example, the terms “social” and “context” have been widely used with assumptions which are not shared. Similarly, the term “community” has been used by different researchers in different ways, and the term “community of practice” has been adopted without regard to the way it has been defined in Wenger’s theoretical framework. There is also a potential danger of appropriating uncritically some of the key notions in other disciplines. For example, the notion of scaffolding might be taken uncritically as assistance which necessarily leads to more effective learning. Another example is the notion of “collaboration” which seems to have been taken as implicated by “interaction.” As Donato (2004) pointed out in his review of current studies of collaborative work, not all forms of classroom interaction are collaborative and conducive to the development of discourse competence.

Another challenge is that the analysis of classroom discourse as situated in its sociohistorical context typically involves an eclectic approach in research methodology and a triangulation of qualitative and quantitative data collected from different sources over a period of time. A rigorous analysis of data requires an iterative process of data interpretation and theory generation which is extremely time consuming. It is sometimes difficult to present a full account of the research processes within the word limit of a journal article (see for example the exemplars of research methodologies presented in Applied Linguistics, Volume 23(3), 2002). This is probably one of the reasons why, as Donato (2004) pointed out, research studies from a sociocultural perspective are rich in theoretical concepts but thin on data.

Yet another challenge is the substantiation of claims made about the relationship between language learning and the classroom discourse data analyzed. For example, claims have been made about the effect of input on learners’ output and the effect of pushed output on language acquisition. However, in many cases, there is a lack of substantial evidence to support such claims. In some cases, the evidence is confined to the learners’ language output in the adjacent discourse units. There is little evidence of the long-term effect of input on language learning. Similarly, claims made about collaborative learning or co-construction of knowledge have been based on the analysis of the co-construction of discourse between the teacher and learners and among learners. While one can argue that the discourse is evidence for co-construction of knowledge, it is not always clear that such co-construction facilitates SL/FL learning. Though some studies have tried to provide evidence of effectiveness on learning outcomes, there appears to be a dearth of large-scale studies to support their educational implications, as Mercer and Dawes (2014) noted. In a review of four decades of research on classroom talk in L1 content classrooms, Howe and Abedin (2013) observed that “much more is known about how classroom dialogue is organized than about whether certain modes of organization are more beneficial than others” (p. 325). The same remark can be made about second and foreign language classrooms. The challenge of establishing the efficacy of certain teacher-student interactions in large scale studies is to ascertain the impact of contextual variables on learning outcomes as the former is not easily controllable in naturalistic settings.

Long (2015) tried to address this issue by proposing conducting research in laboratory settings. He discussed the comparability between laboratory settings and classroom settings and the generalizability of the findings in each. He argued that though results varied with some studies showing greater effect size in one location than the other and vice versa, there was ground for making cautious generalizations about laboratory-based findings to classrooms. He suggested a two-pronged approach to address this issue: If a causal relationship between the areas of under investigation and the learning outcomes is evident in one or more studies in laboratory settings, then classroom studies are warranted. He pointed out that unless there is rigorous empirical evidence to substantiate the claims made about the efficacy of pedagogical strategies, ESL teaching will not be able to establish itself as a profession. This applies not only to the ESL teaching profession but to the entire teaching profession.

Future Directions

Since the 2000s, research on classroom discourse has advanced the field in three aspects. First, there has been a revival of the emphasis on context since the 1990s, an aspect which was minimalized in the 1970s and 1980s. This trend has continued in the past two decades. As we have seen in the review in the preceding sections, instead of focusing on specific aspects of classroom interaction, an increasing number of studies have taken a more holistic view of classroom discourse, attending to the multiple dimensions of context and the multiple levels of discourse in the classroom, relating micro classroom processes to macro institutional and society processes, and engaging in issues such as power, identity, culture, and gender. Studies on the dynamic and dialectical relationship between the processes at these two levels and the agency of the participants in the discourse processes will continue to yield rich insights for what may appear to be mundane routine classroom processes.

Second, the conceptual frameworks drawn on from neighboring disciplines to illuminate the complexity of classroom data have widened considerably and will continue to do so. Van Lier (2000) proposed that the input–output model should be replaced by an ecological perspective. This has been echoed by Kramsch (2002) who considered an ecological approach to language learning as a powerful way of capturing the symbiotic relationship between the language user and the environment. In the collection of chapters in Kramsch (2002), concepts in the sociology of language such as Goffman’s frame analysis and participatory structures have been adopted to analyze the multiple discourse units and levels that are recursively embedded in classroom discourse and the variety of speaker and addressee roles. Papers in the volume point out the need to unravel the cultural, institutional, and interactional dimensions of the contexts in which classroom discourse are embedded. We have seen in this review that increasingly studies of classroom discourse have crossed disciplinary boundaries and have provided a much richer and deeper understanding of classroom discourse. This will continue to be a distinctive feature of research in this field (see also Markee 2015).

Third, a growing number of studies have adopted an eclectic approach to research methodologies in which both qualitative and quantitative data are collected from a variety of sources for triangulation and have provided both etic and emic perspectives in their data analysis. Within the qualitative paradigm, different research methodologies have been drawn on to elucidate the data, as evidenced by the increasing number of studies adopting a conversational analysis approach to the analysis of classroom data.

While classroom discourse research is likely to continue along the trends outlined above, there appears to be five areas which need further work. First, as mentioned before, as the field draws on theoretical concepts and research methodologies from a variety of disciplines, it becomes all the more important that the methodologies and terminologies adopted are explicitly and rigorously defined, with full awareness of their theoretical assumptions, irrespective of whether they have been adopted wholesale, extended, or re-defined.

Second, there has been relatively little in the classroom discourse literature that examines critically the methodological assumptions made in the analysis of data. The special issue in Applied Linguistics (Volume 23(3), 2002), which is devoted to methodological issues in the micro-analysis of classroom discourse, is necessary and timely. It presents a collection of papers containing exemplars and critiques of three influential and well-defined methodologies within which classroom discourse analysis have been conducted: ethnography of communication, conversational analysis, and systemic functional linguistics, which have emerged respectively from anthropology, sociology, and functional linguistics. The discussions do not advocate a particular methodological approach but rather raise researchers’ awareness of methodological issues. More discussion of this kind is necessary to move the field forward.

Third, although there has been an increase in the number of studies adopting both quantitative and qualitative research methodologies, most recent studies are qualitative in orientation. While they have yielded rich insights, it is often not clear whether or not the qualitative findings are peculiar to a specific context. Quantitative data from a larger sample, though not necessarily representative, which can show some general patterns across similar or different contexts would make the qualitative findings much more powerful. Further, the qualitative findings reported are more often than not snapshots of classroom discourse of short durations. More longitudinal studies which show whether there are any changes in the phenomena under investigation are warranted.

Fourth, as noted in the preceding section, the call to move beyond mere descriptions of classroom discourse to establish the relationship between the classroom discourse patterns and processes and learning outcomes should generate more studies that will attend to both the processes as well as products of classroom discourse (Markee 2015).

Finally, the teachers’ and the learners’ voices in the analysis of classroom data continue to be a very important aspect of future research. With the increasing number of studies adopting an emic perspective in the past two decades, the inclusion of the teachers’ voice in the interpretation of data has almost become the norm. The learners’ voice however is still weak. As Cazden (2001) pointed out, classroom discourse should be the object of focal attention for students as well because “all students’ public words become part of the curriculum for their peers” (p. 169). How learners’ engagement in the discourse contributes to the ESL/EFL curriculum constructed in the classroom and how their awareness can be raised are still under-explored.

Cross-References

Related Articles in the Encyclopedia of Language and Education