Teaching Through Dialogue

Inquiry dialogue is initiated by an open, big question about a contestable issue that is relevant to students' lives. The purpose of big questions is neither to test students nor to simply lead them to a narrow range of answers deemed acceptable by the teacher. Rather, these questions invite students to take part in a disciplined inquiry—a higher pedagogical goal. They “problematize, or transform commonly accepted facts or answers into problems to be explored, thereby opening knowledge to thinking” (Lefstein, 2010, p. 176). During inquiry dialogue, students are given considerable control over the flow of discourse: They ask questions, participate in turn management, and evaluate each other's answers. Together, discussion participants search for the most reasonable solution to the contestable question, and if agreement is not possible, they work on clarifying the basis and criteria for their disagreement. The teacher's role is to support collaborative inquiry around contestable questions. During the discussion, teachers “treat students as potential sources of knowledge and opinion, and in so doing complicate expert–novice hierarchies” (Nystrand, Wu, Gamoran, Zeiser, & Long, 2003, p. 140). Importantly, such a view of teacher–student relations does not dismiss the authority of a teacher as a more knowledgeable partner in a discussion. Burbules (1993) has argued that acknowledging authority based on one's expertise or experience does not necessarily threaten the egalitarian nature of interactions and, instead, helps to enhance the intellectual rigor of inquiry.

Below we illustrate the use of inquiry dialogue in a language arts classroom. This excerpt is taken from a recent study that examined teaching practices of elementary school teachers (Reznitskaya, Glina, Carolan, et al., 2012, p. 299). The discussion is based on the story about an unusual giraffe who learns to speak human language and has to decide whether to live with humans or other “regular” giraffes in the zoo (Lipman, 1996). The story prompted students to question what makes one eligible to vote on important decisions.

Note that in this excerpt students and the teacher share control over classroom discourse. The students take over two functions typically performed by the teachers: nominating to speak and, to an extent, questioning. There are exchanges with consecutive student turns without teacher interruption, which demonstrate that students are successfully contributing to the management of turn-taking. The discussion is centered around a big question that is inherently contestable and does not have a clear right answer. As students discuss their positions on the issue of voting, they provide elaborated explanations of reasoning behind their views. The teacher does not dominate the discussion, speaking less than students. He also works to raise the quality of discussion beyond simple sharing of opinions. The teacher's deliberately chosen questions serve to advance the inquiry further, by prompting students to clarify their thinking (i.e., “So, we choose the age to be fair, then?”) or introducing a new perspective, overlooked by the group (i.e., “…and isn't the alternative true?”). Student responses are marked by a high level of collaboration; they agree and disagree with each other's viewpoints, thus developing a more complete and personally meaningful understanding of the issues involved in formulating the required qualifications for voting.

By contrast, very different patterns of engagement can be seen in an excerpt from another classroom. Students in this classroom discuss a Native American tale about a young man who showed kindness to two eagles (Rosebud Yellow Robe, 2001). In return, the eagles saved the man's life by carrying him to safety.

Note that in this excerpt the teacher is clearly the sole authority in the classroom. She controls both the content and the form of the discourse, calling on students to respond, evaluating their answers, and initiating topical shifts with new questions. Her questions largely prompt students to recall the facts from the story. The teacher already knows the correct answers and is the ultimate source of expertise when evaluating students' responses. Her feedback to students represents simple positive reinforcement and the repetition of correct responses. The teacher does not prompt students to support their positions with reasons, to consider alternatives, or to question assumptions and implications. Instead, she moves rapidly from student to student, sacrificing deep intellectual engagement for broad but superficial participation. Students in this classroom do not explain their thinking in depth; their answers are brief, often consisting of only one or two words. They do not collaborate with each other; instead, they direct all their answers to the teacher, rather than peers.

The previous excerpt demonstrates the traditional sequence of (1) teacher question/nomination, (2) student answer, and (3) teacher evaluation, which is often characterized as a “recitation” script. This script is rather common in contemporary classrooms (Applebee, Langer, Nystrand, & Gamoran, 2003; Nystrand, 1997). In a recent observational study of 64 middle- and high-school English classrooms in five states, Applebee et al. (2003) found that the amount of time spent on open discussions averaged only 68 seconds per class. Similar practices have been documented in a large study of teaching practices in the United Kingdom, where the authors routinely observed “the rapid pace of teachers' questioning and the predictable sequence of teacher-led recitation” (Smith, Hardman, Wall, & Mroz, 2004, p. 408). Moreover, recitation script has been shown to be highly resistant to change, even when teachers have progressive notions about the value of dialogue in teaching (Alexander, 2005; Alvermann, O'Brien, & Dillon, 1990). Thus, intensive and sustained professional development is needed to help teachers effectively implement dialogic teaching in their classrooms. We will return to this topic in the second part of the chapter, “Professional Development in Dialogic Teaching.”

Thus far, we explained the concept of dialogic teaching and related terms and used two excerpts to illustrate the key differences between more monologic and more dialogic discussions. However, the picture is more nuanced than these excerpts suggest. Although we privilege inquiry dialogue as the central feature of dialogic teaching, we also recognize the need for other instructional strategies that can be used to support the goals of rational thinking and deeper learning. Alexander (2008) has criticized the use of artificial dichotomies in educational discourse that imply, for example, that teaching students how to think is incongruous with helping them acquire basic knowledge. He explained that teachers should be able to flexibly choose from a broad repertoire of pedagogical approaches, including more traditional methods such as recitation and exposition. For instance, teachers in a more dialogic classroom might, at times, ask students to recall simple facts (i.e., Is there the right to travel in the U.S. Constitution? What does the text say about the safety of self-driving cars?). However, they would do so in the service of promoting well-informed inquiry, rather than simply to have students memorize isolated bits of information. As with any other method, the use of dialogic teaching should not become dogmatic or restrictive. Instead, teachers need to be strategic and flexible in choosing instructional interventions that are consistent with their pedagogical goals, as well as with their assumptions about knowledge, teaching, and learning.

Epistemology and Theory Underlying Dialogic Teaching

We propose that effective use of dialogic teaching requires that teachers develop underlying epistemic and pedagogical frameworks that are aligned with this approach (cf. Kuhn & Udell, 2003; Windschitl, 2002). Models of epistemological development suggest that people progress from a simple view of knowledge as static and known by authorities to a more nuanced understanding of knowledge as socially constructed through the use of reasoning (for review, see Hofer, 2001). Kuhn (1991) offers a useful classification of individual theories of knowledge, proposing three stages of development: absolutist, multiplist, and evaluatist. Absolutists view knowledge as fixed, certain, and existing independently of human cognition. Multiplists see knowledge as entirely subjective, denying the role of reason and expertise and considering all opinions to be equally valid. At the most advanced stage, evaluatists accept the subjective nature of knowledge, while also recognizing that we can engage in a rational evaluation of different viewpoints and, as a result, consider some judgments to be more reasonable than others.

Absolutist and multiplist epistemologies are incompatible with dialogic teaching. Teachers at the absolutist level believe that only authority figures have legitimate knowledge, so there is no need to involve students in knowledge co-construction, critique, and evaluation. Teachers with multiplist views fail to appreciate the value of dialogue because they see knowledge as entirely relative and idiosyncratic, discounting the use of reasoning to support and justify claims. By contrast, teachers who subscribe to an evaluatist epistemology are more likely to successfully implement dialogic teaching because they view knowledge as “the product of a continuing process of examination, comparison, evaluation, and judgment of different, sometimes competing, explanations and perspectives” (Kuhn, 1991, p. 202).

In addition to evaluatist epistemology, dialogic teaching reflects a social-constructivist perspective on learning (e.g., Piaget & Inhelder, 1969; Vygotsky, 1968). According to this perspective, students are active meaning-makers who learn through constructing and negotiating new understandings in interaction-rich communities of practice. Language is viewed as not only a medium for communicating ideas, but also as a primary tool for forming new ways of thinking and knowing (Vygotsky, 1968). Following Vygotsky, we believe that argument skills are developed, at least in part, through participation in a collective activity, such as a discussion about a contestable question during which students experience novel language and thought practices. Although each discussion is different, participants engage with common “cultural tools” (Vygotsky, 1981) characteristic of argumentation, including taking a public position on an issue, supporting it with reasons and evidence, challenging other participants, and responding to counterarguments with rebuttals. Over time, these processes, which are made “visible” in a group's argumentation, become part of an individual's repertoire of thinking strategies. Thus, collaborative engagement in inquiry dialogue represents a useful model and a training ground for the development of argument skills. As participants in discussions collectively formulate, defend, and scrutinize each other's viewpoints, they begin to internalize general intellectual dispositions, skills, and knowledge of reasoned argumentation.

To further define the learning outcomes resulting from student engagement in inquiry dialogue, we rely on constructivist approaches, in particular schema theory (e.g., Anderson, 1977; Rumelhart & Ortony, 1977). Schema theory proposes that knowledge can be represented as organized mental structures, or schemas. Learning involves generation and modification of these schemas, and successful transfer entails accessing and using relevant abstract structures (Gick & Holyoak, 1987; Reed, 1993). Applying schema theory to argumentation development, we propose that, through consistent engagement in inquiry dialogue, students come to recognize important commonalities in their experiences and, as a result, develop an internal abstract knowledge structure we call an argument schema (Reznitskaya et al., 2008). In order to specify the elements of an argument schema, we draw upon the normative models proposed by argumentation scholars (e.g., Toulmin, 1958; Walton, 1996). Specifically, we suggest that a well-developed argument schema includes such elements as position, reason, evidence, warrant, counterargument, and rebuttal. It contains the knowledge about the criteria for argument evaluation (i.e., validity of inferences, acceptability of premises), as well as the informal heuristics of reasoning. This knowledge supports comprehension, construction, and evaluation of arguments in new contexts. Importantly, different components of an argument schema are linked through a higher-order mental structure that “glues together” pieces of information that otherwise would remain unrelated or acausal (Mishra & Brewer, 2003). Following Kuhn (1999), a developed argument schema is supported by epistemic beliefs that recognize the function and value of a rational argument as a means for choosing among alternative propositions or actions, or an evaluatist epistemology.

To recap, in a more dialogic classroom, students experience new ways of formulating and expressing their viewpoints. As students discuss big, contestable questions with their peers, they observe, practice, and eventually internalize the skills of argumentation into individual mental structures, or argument schemas. Students can then use their argument schemas in other contexts, such as when discussing a new issue or when reading and writing arguments. Supporting and informing the entire process of argumentation development is a progressive shift in teachers' and, eventually, their students' epistemologies from an absolutist or multiplist view of knowledge to an evaluatist view. Teachers and students who subscribe to an evaluatist epistemology are more likely to view and use inquiry dialogue as a means of forming and evaluating different points of view. In a similar way, Kuhn (1999) suggests that advanced levels of epistemology are essential for engagement in argumentation, as they provide reasons for actually using the skills of argument when solving ill-structured problems. Consequently, evaluatist epistemology is both the necessary context for dialogic teaching and an important learning outcome for the students.

Empirical Research on Dialogic Teaching

When examining research studies on dialogic teaching (e.g., Alexander, 2003; Applebee et al., 2003; Billings & Fitzgerald, 2002; Mercer & Littleton, 2007; Soter et al., 2008), we can see considerable consensus as to what constitutes typical features of more dialogic classrooms. Supporting the theoretical propositions outlined earlier, research reveals that dialogic classrooms have more egalitarian social organization, with authority over the content and form of discourse shared among discussion participants. Student learning is supported by open, cognitively challenging questions that engage students in critical evaluation and analysis. During the discussions, teachers play an important role by facilitating focused and rigorous inquiry: They build upon student answers, ask for clarification, prompt for alternative perspectives, and encourage students to relate their ideas to those of their peers in the discussion. As a result, students in more dialogic classrooms have new opportunities to engage in collaborative construction of knowledge, because they listen to and react to each other's positions and justifications.

We now turn to studies that examined the learning processes and outcomes in a more dialogic classroom. Although such research is still tentative, studies have generally documented that, following engagement in discussions, students show gains in reasoning in new contexts, argumentative writing, inferential comprehension of text, as well as deeper conceptual understanding of disciplinary concepts and principles (Asterhan & Schwarz, 2007; Dong, Anderson, Li, & Kim, 2008; Kuhn & Crowell, 2011; Mercer, Wegerif, & Dawes, 1999; Reznitskaya et al., 2001). For example, students' ability to generate arguments that contained more than one perspective improved as a result of their participation in dialogic inquiry (Kuhn & Crowell, 2011). Similarly, in a review of several quasi-experimental studies on dialogic teaching (Reznitskaya et al., 2009), the authors concluded that students who participated in discussions about their readings consistently wrote essays that contained a greater number of argument components, including supporting reasons, counterarguments, and rebuttals, compared to their peers from control classrooms. The authors suggested that the experience of engaging in inquiry dialogue with peers allowed students to internalize important knowledge of argumentation. For example, it helped students to consider and integrate alternative positions in their compositions, thus supporting an important shift from monologic to dialogic thinking.

There is also emerging evidence regarding the influential role of personal epistemologies in teaching and learning. Studies of teacher epistemology have shown that beliefs about knowledge and knowing are often consistent with observed classroom practices (e.g., Schraw & Olafson, 2002; Sinatra & Kardash, 2004; Stipek, Givvin, Salmon, & MacGyvers, 2001). For example, in a study by Johnston, Woodside-Jiron, and Day (2001), researchers found that teachers' epistemologies were directly aligned with their instruction, influencing the power relations between teachers and students and their interactional patterns, including the type of questions discussed and the feedback given to students. However, the relationship between teacher beliefs and practice is not simple. Subscribing to more sophisticated ideas about knowledge and knowing might not always translate into the use of inquiry dialogue in a classroom (Alvermann et al., 1990; Schraw & Olafson, 2002), especially when teachers are in the processes of changing their beliefs, with “changes in beliefs preceding changes in practice” (Richardson, Anders, Tidwell, & Lloyd, 1991, p. 579).

Research on learner epistemologies has revealed that students at more advanced stages are more likely to better comprehend texts, to develop a deeper conceptual understanding of a given subject, to identify informal reasoning fallacies, and to construct arguments of higher quality (e.g., Mason & Scirica, 2006; Qian & Alvermann, 2000; Stromso & Braten, 2009; Weinstock, 2006; Weinstock, Neuman, & Tabak, 2004). These findings are consistent with the theoretical claims outlined earlier in this chapter, indicating that epistemological beliefs may act as general filters that direct one's engagement with argumentation.

In sum, argumentation skills are considered to be of vital importance for living a productive and meaningful life in a postindustrial globalized world of the 21st century. According to contemporary theory and research, these skills can be developed through dialogic teaching—a pedagogy that strategically uses classroom talk to promote students' thinking. Yet, studies continue to document that dialogic practices are rare in today's classrooms (Alexander, 2005; Mehan, 1998; Nystrand et al., 2003; Smith et al., 2004). Considering the consistent findings from research on classroom discourse, it is possible to imagine that many students go through schooling without ever experiencing a genuine discussion during which they are invited to collaboratively reason through a complex question in a systematic and rigorous way.

In an effort to address the apparent disparity between the higher educational goal of teaching argumentation and the reality of typical classroom practices, we have begun to design a comprehensive professional development program in dialogic teaching. This program targets elementary school teachers in language arts classrooms. We now turn to the discussion of the development and implementation of our program.

Overview of the Professional Development Program

In each year, we are trialing a version of the program and collecting data from teachers and students to assess program effectiveness and inform its revisions. In other words, each year comprises a new iteration of the program. The program has the following goals:

• To help teachers develop the pedagogical and epistemic views that are aligned with dialogic teaching.
• To sensitize teachers to the nature of talk and inquiry dialogue.
• To help teachers understand argumentation and criteria for evaluating arguments (i.e., develop their argument schemas).
• To provide teachers with a repertoire of moves to facilitate inquiry dialogue.

At the time of writing, we have completed the first year of the project. We conducted the project as a design study (Collins, Joseph, & Bielaczyc, 2004), during which we worked collaboratively with teachers throughout the year to identify and organize instructional activities that support teachers' knowledge and use of dialogic teaching to promote argumentation.

Participants

In year 1, we worked with 10 grade 5 teachers and their students. Study participants came from school districts in two states, Ohio and New Jersey. There were six teachers at the Ohio site and four teachers at the New Jersey site, with teaching experience ranging from 2 to 22 years.

Procedure

The project in year 1 was conducted in three stages: pretesting, implementation of the professional development program, and posttesting. All project activities were similar at both sites, but with some variations to test the viability and effects of different instructional approaches or sequences of approaches. During pretesting stage (September), we video-recorded two discussions in each classroom in order to collect baseline information about typical teacher practices. We also interviewed teachers about their background and experience, as well as assessed their epistemological beliefs using the interview measure developed and validated by King and Kitchener (1994).

Starting in October, we began the implementation of the professional development program. We asked teachers to conduct discussions with their students at least once per month during their language arts lessons, and we video-recorded these discussions. We also met with teachers every 2 weeks in teacher study groups. Study groups lasted for about 1.5–2 hours, totaling approximately 30 hours per year at each site. During these study-group meetings, participating teachers engaged in mini-lessons, group activities, collaborative analysis of videos of classroom discussions, and exercises on topics related to dialogic teaching, inquiry dialogue, and argumentation. In addition, teachers read and discussed articles and chapters on these topics (e.g., Govier, 2010; Michaels, O'Connor, & Resnick, 2008; Waggoner, Chinn, Yi, & Anderson, 1995). After every four study group meetings (i.e., approximately every 2 months), we conducted focus-group interviews with teachers in each study group. The purpose of these focus-group interviews was to identify what teachers found valuable (or not) in learning about dialogic teaching and argumentation. Teachers also received individual coaching in how to conduct discussions to promote argumentation. During these sessions, teachers viewed and critiqued their own classroom interactions with the help of an experienced discourse coach, who supported the teachers' ongoing development and reflection. All study group meetings, focus-group interviews, and coaching sessions were audio-recorded.

Throughout the implementation of the program, teachers also practiced using three existing classroom observation tools to assess talk in their video-recorded discussions: the Instructional Quality Assessment Tool (Junker et al., 2006), the Talk Assessment Tool for Teachers (Wilkinson, Reninger, & Soter, 2010), and the Dialogic Inquiry Tool (Reznitskaya, Glina, & Oyler, 2012). Finally, teachers routinely documented reflections about their learning by posting to a discussion board on a program website.

During the posttesting stage (May), we again interviewed teachers about their epistemological beliefs. We also piloted student measures designed to assess argument skills when speaking, reading, and writing. These measures will be used to evaluate program effectiveness in years 2 and 3.

Curriculum Materials and Data Sources

At the end of year 1, we developed initial materials for the program, including PowerPoint slides, instructional activities, videos for illustrating inquiry dialogue, and readings for teachers and students. We also collected data from multiple sources, including study-group meetings, focus-group interviews, coaching sessions, and teacher postings to a course website. We are conducting content analysis of the data to inform the revisions of the professional development program. In the next section, we discuss lessons learned as a result of year 1 implementation.

Lessons Learned

The design of our program was informed by the principles of effective professional development described in previous studies (e.g., Dole, 2003; Elmore, 2002; Hawley & Valli, 1999; Wei, Darling-Hammond, Andree, Richardson, & Orphanos, 2009; Yoon, Duncan, Lee, Scarloss, & Shapley, 2007). These principles include: the involvement of teachers in planning and undertaking professional development, the collective participation of groups of teachers, intensive and sustained support, and the integration of professional development into the daily lives of teachers. In year 1, we were able to translate these design principles into specific instructional activities. To address the need for collegial and ongoing engagement that is relevant to classroom practice, we structured our program as a series of 12 biweekly study-group meetings, followed up by in-class video-recording and individual coaching. During the study group meetings, teachers collaborated with each other and the researchers to learn about dialogic teaching, inquiry dialogue, and argumentation.

One helpful activity that emerged from the study-group meetings was coplanning. During coplanning, teachers or researchers brought examples of children's stories and articles to the meetings. Next, we selected a text for discussion and identified important themes from this text. For instance, an article about a boy that was paralyzed after getting a concussion during a football game brought up such themes as sports violence, the role of society in encouraging aggressive behaviors, the differences in responsibilities of adults versus children, groups versus individuals, and so on. These themes were then turned into big questions that served to launch the discussion. Next, study-group participants practiced engaging in inquiry dialogue around a contestable question (e.g., Who was responsible for the boy's injury?). The discourse coach acted as a discussion facilitator, modeling good practice for the teachers. After spending time on deliberating the big question, teachers evaluated the quality of their own discussion by reflecting on helpful discourse moves that supported argumentation.

From observations and teacher feedback, we learned that coplanning and engaging in inquiry dialogue provided a useful means for learning about various aspects of inquiry dialogue, including: (a) identifying contestable issues in children's literature, (b) launching the inquiry with a big question, (c) supporting argumentation development during the discussion with strategic facilitation moves, and (d) evaluating the processes and products of the discussion. As teachers discussed contestable questions with others, they were able to personally experience the possibilities and challenges of being a participant in an inquiry dialogue: They tried out new participation structures, developed and clarified their thinking on a big question, and had their ideas challenged by peers. We believe that these experiences helped teachers to acquire procedural knowledge about facilitating inquiry dialogue and to connect it to conceptual knowledge about the role of language in student learning. Coplanning and engaging in inquiry dialogue also offered teachers a shared experience that they were able to take back to the classroom, thus grounding their learning in practice.

In addition, teachers received ongoing support and targeted feedback about their implementation of dialogic teaching through individual coaching. During the coaching sessions, a teacher and a coach watched a recording of a recent discussion and engaged in collaborative problem solving and reflection. Teachers' reactions to coaching were uniformly positive, as illustrated by the comments below:

• “The one-on-one coaching I found to be, I think, the most helpful only because it's so hyperspecific.”
• “I thought [the coaching] to be the most helpful because it was even better than just watching the video…it was watching it with purpose.”
• “The coaching has been phenomenal. Watching the video with a coach and getting feedback on how I could have made a different move has helped me transfer that right into my teaching. It has also helped when I would watch by myself and feel like something had gone wrong, but didn't know what to do in the moment. Then I would show that to the coach and they would help me figure out what to do next time.”
• “A couple of things [the coach] brought up that I just simply had never even considered…that he called out and said, ‘You maybe could have done this here.’”

It appears that teachers appreciated having an opportunity to experiment with new dialogic practices in their classrooms, to reflect on the quality of their talk, and to evaluate their progress with an expert colleague in a systematic and supportive manner.

Another principle from research on professional development highlights the importance of having practitioners develop coherent pedagogical frameworks that integrate both theoretical and practical knowledge (e.g., Elmore, 2002). Real transformation in teaching happens only when teachers

think differently about what is going on in their classrooms, and are provided with the practices that match the different ways of thinking. The provision of practices without theory may lead to misimplementation or no implementation at all… [C]hanging beliefs without proposing practices that embody those theories may lead to frustration. (Richardson et al., 1991, p. 579)

One of our challenges in year 1 was to find the right balance between theoretical and practical knowledge germane to dialogic teaching and to develop productive ways of teaching both. For example, our model of dialogic teaching assumes that evaluatist epistemology is a normative framework for effective facilitation of inquiry dialogue. In other words, teachers need to develop underlying assumptions about the nature of knowledge and the process of knowing that are aligned with dialogic teaching.

Unfortunately, there are very few studies that have examined pedagogical interventions that can bring about changes in epistemological beliefs of teachers (Brownlee, Purdie, & Boulton-Lewis, 2001; Hill, 2000). This is especially problematic, considering that college education, including teacher preparation programs, is not successful at helping learners advance their epistemologies to the evaluatist level (Brownlee et al., 2001; Schraw & Olafson, 2002). Hofer (2001) has concluded her review of research on personal epistemologies by lamenting that “our ‘educated citizenry’ may in fact be largely composed of individuals who view the world from a position of absolutism, or who simply accept a multiplicity of opinions about complex issues, seeing no need to support positions with evidence” (p. 369). Our own findings in year 1 confirm Hofer's assessment. Using epistemology interviews developed by King and Kitchener (1994), we found that none of the teachers started the program at the evaluatist level.

One potentially effective way to support teacher epistemological development that emerged from year 1 implementation is to involve teachers in inquiry dialogue about knowledge construction itself. For example, during one study-group meeting, teachers collaboratively deliberated about the question “What makes for a better answer during any group discussion?” We observed the teachers struggling with identifying the criteria for “a better answer,” often asserting that “everyone is entitled to their own opinion.” Although we hoped teachers would be able to move beyond this multiplist position, we also realized that a group inquiry into the questions of knowledge and knowing was highly effective for uncovering teachers' views and unexamined assumptions, thus making them available for analysis and reflection. We plan to expand on the idea of using collaborative discussions about epistemology during year 2 implementation, as it has also been shown to be effective in prior research (e.g., Hill, 2000).

In addition to identifying ways to address core theoretical issues, we also struggled with helping teachers understand how abstract theoretical principles of dialogic teaching can be transformed into specific classroom practices. As mentioned earlier, there is considerable consensus about what is involved in conducting productive discussions about text (e.g., Wilkinson, Soter, & Murphy, 2010). Yet, learning to implement such practices presents a serious challenge for teachers (Juzwik, Sherry, Caughlan, Heintz, & Borsheim-Black, 2012). Both novice and experienced teachers have difficulties facilitating classroom discussions in a way that supports argumentation development (e.g., Alvermann & Hayes, 1989; Caughlan, 2003; Chinn, Anderson, & Waggoner, 2001). We believe that one reason for these difficulties has to do with insufficient opportunities teachers have to learn about the structure and the criteria for evaluating arguments. As a result, teachers may not be able to identify faulty reasoning, spot contradictions in students' claims, or hear unwarranted conclusions. Consequently, teachers might miss opportunities to make effective facilitation moves. In the absence of teacher intervention, the discussion is likely to disintegrate into simple sharing of opinions, which remain disconnected and unexamined.

We discovered that an effective approach to introducing teachers to the essentials of argument and argumentation is to situate this learning in the context of a classroom discussion. For example, during study-group meetings, we used videos and transcripts of classroom discussions to have teachers practice tracking and evaluating group arguments. As teachers watched the video or read a transcript of a group discussion, they worked on: (a) identifying key elements of an argument (i.e., positions, reasons, evidence, warrants, etc.), (b) reflecting on the quality of argumentation, and (c) suggesting effective facilitation moves to improve students' reasoning. For instance, when student arguments lacked clarity, a teacher might intervene with clarifying, summarizing, or revoicing moves, such as “What do you mean by…? So, are you saying that…? How are you using the word…?” Similarly, to test the validity of inferences in an argument, the teacher might ask questions such as: “Does it follow? Is that the only explanation? How is this relevant?” We have learned that teaching about arguments needs to be connected to facilitation moves. There seems to be little sense in describing teachers' moves in isolation from student arguments during the discussion, and vice versa.