Socializing Intelligence Through Academic Talk and Dialogue

The American Educational Research Association (AERA) publishes books and journals based on the highest standards of professional review to ensure their quality, accuracy, and objectivity. Findings and conclusions in publications are those of the authors and do not reflect the position or policies of the Association, its Council, or its officers.

© 2015 American Educational Research Association

The AERA Books Editorial Board

Chair: Gilberto Q. Conchas

Members: D. Jean Clandinin, Jeffrey R. Henig, Felice J. Levine, Simon W. Marginson, Nailah Suad Nasir, Charles M. Payne, Russell W. Rumberger, Mariana Souto-Manning

Published by the American Educational Research Association

1430 K St., NW, Suite 1200

Washington, DC 20005

Printed in the United States of America

All rights reserved. No part of this publication may be reproduced or distributed in any form or by any means, including, but not limited to, the process of scanning and digitization, or stored in a database or retrieval system, without the prior written permission of the publisher.

Library of Congress Cataloging-in-Publication Data

Socializing intelligence through academic talk and dialogue / edited by Lauren B. Resnick, Christa Asterhan, and Sherice N. Clarke.

pages cm

Includes bibliographical references and index.

ISBN 978-0-935302-70-7 (pbk. : alk. paper) -- ISBN 978-0-935302-40-0 (hardback : alk. paper) -- ISBN 978-0-935302-43-1 (e-book) -- ISBN 978-0-935302-62-2 (kindle) -- ISBN 978-0-935302-61-5 (pdf)

1. Interaction analysis in education. 2. Communication in education. 3. Student-teacher relationship. 4. Dialogue. 5. Discussion 6. Active learning. I. Resnick, Lauren B., editor of compilation. II. Asterhan, Christa, editor of compilation. III. Clarke, Sherice N., editor of compilation.

LB1034.S635 2015

371.102’3--dc23

2015007798

Acknowledgments

Many thanks to the American Educational Research Association for awarding a special grant to support the AERA Education Research Conference that led to this volume. We also thank the Learning Research and Development Center and the Pittsburgh Science of Learning Center for providing additional funding for the research conference. Special thanks to Deanna Weber Prine, who oversaw our internal editorial process, and Faith Schantz, and also to the AERA publications department staff, John Neikirk, Martha Yager, and Jessica Campbell, who so ably brought the project to completion under a tight timeline.

Table of Contents

I N T R O D U C T I O N

Talk, Learning, and Teaching

LAUREN B. RESNICK

Learning Research and Development Center, University of Pittsburgh

CHRISTA S. C. ASTERHAN

Hebrew University of Jerusalem

SHERICE N. CLARKE

Learning Research and Development Center, University of Pittsburgh

In September 2011, the American Educational Research Association sponsored a research conference in Pittsburgh that brought together leading scholars from across the world in education, learning sciences, cognitive psychology, educational psychology, linguistics, and computer science. During the previous two years, we had communicated with scholars from many countries who we knew were interested in the roles of discussion and social interaction in school learning. We asked them to send us any data they had (even if it was still unpublished) on the effects of carefully orchestrated discussions on student learning. We also asked them to suggest other scholars to whom we might write.

The responses were startling. The data included evidence that students who had experienced this kind of structured dialogic teaching performed better on standardized tests (i.e., tests that the investigators did not control) than similar students who did not have discussion experience. The data also showed that some students retained their learned knowledge for two or three years. More surprising, in some cases students even transferred their academic advantage to a different domain (e.g., from science instruction to an English literature exam). These results were not found every time teachers tried to use dialogic methods in teaching traditional subject matter. But they occurred with enough frequency, and in enough of a variety of countries and school environments, that we began to think we had happened upon a powerful new way of organizing school learning. We saw the potential for providing educators with reliable tools for more equitable ways of teaching and students with transferable skills and knowledge.

Of course, we (and the scholars and educators who sent us their work) had not just happened to think of classroom talk as a way of expanding educational achievement and opportunity. All of us were responding to a changing landscape of education practice and social theory that had been developing for more than a century.

Over the course of the 20th century, and continuing into the present, virtually every nation in the world has expanded its educational aspirations. Who went to school, what they learned, and what they could do with their knowledge beyond the classroom became a central concern. More children now go to school, and for longer periods, than at any time in the past. They are expected to acquire more complex knowledge, along with skills for using that knowledge. Such learning is believed to both create personal opportunity and contribute to society. As the population of schools has changed and the needs of democratic societies have become more evident, educators and others have voiced an active concern about equity, both in participation and in the outcomes of schooling. The demand to teach a greater body of knowledge to more students has led educators to experiment with new ways of organizing schools and teaching within them.

All of this has occurred against a backdrop of major changes in the social, cognitive, and learning science disciplines that form the basis for educational research:

The idea that knowledge and skills are nothing more than collections of multiple bits of information—a view that predominated in learning and cognitive psychology for decades—is now giving way to complex (and widely debated) theories of how the bits of information are organized and used.

Even scholars who work within classical cognitive theory, with a focus on individual competence and performance, have recognized that social processes and social expectations provide frameworks within which individuals learn (Koedinger, Corbett, & Perfetti, 2012).

It is now understood that knowledge develops and is constructed through social interaction, including debates and disagreements in various disciplines (Bakhtin, 1981; Dewey, 1916; Doise, Mugny, & Perret-Clermont, 1975; Mead, 1967; Resnick, Levine, & Teasley, 1991; Vygotsky, 1962, 1978), although we still do not know exactly what kinds of exchanges are most effective.

We understand language and linguistic competence in new ways, including differences in privileges for speaking (Bourdieu, 1991; Gee, 1989) and styles of language (Heath, 1983). We know that multiple linguistic forms can be used to express intended meaning and complex ideas, to coordinate actions (Resnick, Levine, & Teasley, 1991; Resnick, Säljö, Pontecorvo, & Burge, 1997), and to mark one’s belonging to certain social groups.

The concept of culture has become a major element of educational and social research. Researchers examine how schools adapt (often negatively, occasionally positively) to cultural differences among students. Ladson-Billings, among others, has shown that these concerns can be positively addressed within a high-demand intellectual environment (2009).

The social design of classrooms has been shown to influence students’ achievement and their retention of knowledge (Dweck, 2006; Steele, 1997).

In response to these theoretical shifts, new theories of instruction have arisen. Rather than just having students memorize and recite material presented by the teacher or by a written text, many kinds of discussion-based teaching methods have been proposed, and a few have been implemented and studied.

With these developments as background, we came together in Pittsburgh to talk about talk—specifically, the role of academic dialogue in learning. Working from different theoretical perspectives and research traditions, the scholars who attended the conference have generated new inquiries and approaches to the study of talk and produced some surprisingly powerful evidence. However, given the differences in our methods of study, it has been difficult to cumulate the knowledge, identify disagreements, and design research and practice efforts that have a good chance of succeeding. Prior to the Pittsburgh meeting, the work took the form of islands of evidence, unconnected to a whole. We organized the conference as a working meeting to interrogate the boundaries of our respective islands. Its primary goal was to place these varied bodies of research and evidence in conversation with one another. In this volume, we attempt to capture some of the excitement of the Pittsburgh conference, and to indicate possible directions for continuing research.

The conference was not just an academic undertaking, however. The collection of scholarship offered here has an embedded social mission. This book offers evidence that certain kinds of structured discussions can produce enduring learning and that, under the right conditions, virtually every student can participate. The kind of classroom talk that our authors investigated accepts students’ emergent ideas regardless of whether they are framed in proper speech—no grammar or vocabulary test is required to participate. In this sense, dialogic teaching has the power to break the cycle of low demand/low performance too often experienced by children from disadvantaged socioeconomic backgrounds, children who are ethnic minorities, and/or those who are not fluent in the dominant language. All students are invited into discussions that allow all (respectful) forms of expression. This form of talk increases cognitive demand, and hence learning opportunities. Students do not just chatter. They must defend their statements—for example, by referring to a text, demonstrating a mathematical proof, or citing scientific evidence. In short, students must make a claim on truth.

Overview of the Volume

This volume assembles the edited papers, and some of the discussion, that emerged from the conference. We bring together a wide range of scholarship on dialogue, and its disparate nature is reflected in the use of terminology. Authors use dialogue, dialogic pedagogy, dialogic teaching, accountable talk, deliberation, and argumentation, but all share basic assumptions. For the authors, talk is a privileged form of learning. This kind of talk begins with students thinking out loud about a domain concept: noticing something about a problem, puzzling through a surprising finding, or articulating, explaining, and reflecting upon their own reasoning. Students do not simply report facts they already know for the teacher to evaluate. Instead, with teacher guidance, they make public their half-formed ideas, questions, and nascent explanations. Other students take up their classmates’ statements: challenging or clarifying a claim, adding their own questions, reasoning about a proposed solution, or offering a counterclaim or an alternate explanation. This form of talk is orchestrated by a teacher. It may be conducted in whole groups, smaller collaborative groups, or with pairs of students. The key component is the learning power generated by two or more minds working on the same problem together.

Following a Prologue that invites readers into one of the earliest systematic studies of classroom talk, we offer five sets of chapters. Each section examines questions that participants considered at the Pittsburgh meeting and questions that emerged in light of the discussion. Two commentary chapters at the end reflect on the themes and data presented throughout the book and suggest directions that continuing research and practical development efforts might fruitfully explore.

Section 1 examines the evidence on learning through structured discussion. Section 2 examines the nature of productive talk across a range of learning settings (informal and formal). Section 3 considers dialogue in computer-mediated environments. Section 4 considers theoretical explanations that can account for the effects of discussion-based instruction, and methodological approaches to examining classroom talk in the age of big data. Section 5 examines how such instruction might be brought to scale through teacher professional development. Finally, the two chapters of the Epilogue reflect on the range of issues that have been considered and suggest some profitable next efforts for the field.

Prologue: Problematizing Classroom Talk

We open the volume with a Prologue in which authors Hugh Bud Mehan and Courtney Cazden reflect on their seminal study that set this field in motion nearly 40 years ago, and relate it to present research. Cazden, on a sabbatical from Harvard University, returned to the classroom for a year to teach an elementary school class. Mehan, a sociologist equipped with the new tools of microsociology and ethnomethodology, studied her practice. The result was pioneering research that described a basic structure of classroom life: The teacher initiated (I) the talk, often by posing factual questions; students responded (R) briefly; and the teacher evaluated (E) the response before moving on to a new question.

Mehan and Cazden also documented classroom discourse beyond the three-turn IRE exchange for the first time. Traditional information, such as arithmetic, was taught mainly through IRE recitation, but Cazden found she could open a new conversational space during social studies lessons, in which she asked students questions about their families—questions to which she did not know the answers. With this shift in authority, she saw the classroom come to life as a social place. Many of the later chapters in this volume can be thought of as explorations of the ways in which Cazden’s methods of opening conversation can also be used to teach the core content of the school curriculum.

Section 1: Effects of Dialogic Participation in and Beyond the Classroom

The assembled chapters making up Section 1 document student growth in the ability to participate in academic discussions; students’ mastery of traditional subject matter knowledge in the core domains of mathematics, science, and language arts; and their ability to apply learning in other domains (transfer). The chapters in this section can be read in any order. Taken together, these chapters show that dialogic forms of teaching can significantly raise student achievement—on traditional tests and examinations that educators have relied on for decades—in virtually every school subject. Several of the chapters also show strong retention of knowledge over the years, along with transfer to other school disciplines.

In Chapter 3, Ian Wilkinson, Karen Murphy, and Sevda Binici discuss the extensive research literature on discussion-based teaching of reading comprehension. They review what is known about the major approaches for conducting discussions about text and show that structured book discussion can produce important reading comprehension gains. They also point out that only a few controlled studies have looked at whether discussion fosters general comprehension abilities, that is, whether these abilities transfer as individuals engage with new texts and novel tasks. Even though more controlled research on transfer is needed, the authors’ review of the best evidence available suggests that students may indeed be learning general text interpretation skills and perhaps developing enthusiasm for engaging with complex written material. In other words, they appear to be learning how to build interpretations—not just acquiring knowledge of specific texts or vocabularies.

The next two chapters are written by authors whose reading programs are included in Wilkinson and colleagues’ list of programs that have documented effects on reading competence. Margaret McKeown and Isabel Beck (Chapter 4) provide a detailed account of how their Questioning the Author program guides elementary school learners through an imagined conversation in which they press the author to explain why his or her text used particular words, phrases, and text structures. As the chapter title suggests, comprehending a text is viewed as an interactive process, one in which both author and readers have a voice.

In several important studies, Richard Anderson and his colleagues have documented growth in reading comprehension test scores and argumentative reasoning as a function of the semistructured classroom discussion of children’s stories (e.g., Reznitskaya et al., 2009). In Chapter 5, Jingjing Sun, Anderson, Tzu-Jung Lin, and Joshua Morris attempt to open the black box of learning and development that occurred across a successive series of group discussion activities. Adopting a microgenetic approach, they carefully trace the emergence of elementary school students’ learning to lead discussions and learning to apply analogic reasoning in peer dialogue.

Deanna Kuhn and Nicole Zillmer (Chapter 6) show a similar trajectory of growth in the ability of middle school students to engage in reasoned discussion as a result of participating in student-led discussions twice weekly over three years. Chapters 5 and 6 might be described as accounts of programs in which students are placed in small groups and let loose to develop reasoning skills by arguing with one other. However, a careful reading of each will show that the sequence of texts and topics for discussion, together with some ground rules introduced and monitored by adults, are crucial for the development of student skills.

Noreen Webb, Megan Franke, Angela Turrou, and Marsha Ing (Chapter 7) show how teacher modeling and carefully orchestrated teacher interventions are essential to children’s learning elementary mathematics during group discussions. In particular, Webb et al. help to identify precisely the kind of teacher support that improves students’ collaborative problem solving, and the kind of intervention that is less effective.

In the final three chapters of this section, we encounter evidence that students can learn more than just what they are taught when teacher-led discussion is carefully crafted. These chapters document what some have called the holy grail of education: retention of intellectual advantage over time, and transfer to disciplines and skill sets that were not directly taught.

Keith Topping and Steven Trickey (Chapter 8) show that one hour of discussion per week of texts developed to teach basic philosophical concepts to children led to superior performance on the Cognitive Abilities Test, a test that measures nonverbal and quantitative reasoning as well as verbal reasoning skills. Moreover, these advantages over a control group were retained two years after the experimental program ended and after students had transferred from primary to secondary school.

Catherine O’Connor, Sarah Michaels, and Suzanne Chapin (Chapter 9) describe an intervention that included structured discussion in elementary and middle grade mathematics classrooms. Students in the intervention group, from families of very low socioeconomic status, performed at high levels on the state English language arts test as well as on the state mathematics test.

A skeptic might consider each of these smaller scale studies as statistical flukes—unlikely to be repeated at a scale of interest to education designers or policy makers. However, the work of Philip Adey and Michael Shayer (Chapter 10) over a 30-year period shows that such findings can occur at a scale that scholars and educators cannot ignore. Adey and Shayer studied the effects of structured, teacher-led discussion of science and mathematics problems (mostly based on the interview protocols of Jean Piaget and his colleagues in Geneva) on long-term mental development. Their most impressive set of findings shows that participants in the science discussions at 12 years of age outperformed control groups three years later, not only on the science portion of the British national examinations but also on the English and mathematics portions.

Section 2: Dialogic Classroom Cultures

The chapters in the second section take an in-depth look at several examples of academic dialogic instruction. The authors assume, in effect, that dialogic participation can produce cognitive gains (as shown in Section 1). They ask, What does it look like? and How are academically productive discussions facilitated? within and outside of the classroom. Making discussion and dialogue an integral part of learning in either type of setting poses pedagogical as well as epistemic challenges, many of which are addressed here. These chapters focus on the tension between (a) the roles of the teacher as a discussion leader and as an authority in the classroom, (b) registers of knowledge and discipline-specific ways of knowing, and (c) student autonomy and student voice within the classroom.

In Chapter 11, Michael Ford and Ellice Forman show that most students think of science as a body of known facts rather than a continuously evolving set of explanations. The authors use Bakhtin’s concept of dialogic thinking to describe the process of creating scientific explanations—showing how scientists actively seek data and interpretations that challenge current scientific thinking. Ford and Forman show how a skillful teacher can gradually shape a community of discourse in which students learn to accept, even to search for, uncertainties that can ultimately be used to build authoritative scientific knowledge.

In Chapter 12, Baruch Schwarz takes us inside a learning environment that has rarely been observed directly by outsiders—the discussion of biblical and Talmudic texts by students in a traditional Jewish yeshiva. Following a tradition developed over many centuries, young men read and discuss texts, rabbinic commentaries from different time periods, and textual interpretations that have become the basis of written and oral law. In yeshiva debates, study partners develop and critique each other’s interpretations according to standards of analysis and dialectical argumentation. Schwarz suggests that the rules of yeshiva partner study and those of scientific argumentation are surprisingly similar. He raises the question of whether yeshiva-style argumentation might be adopted fruitfully in science teaching.

In Chapter 13, Sherice Clarke shines a spotlight on students’ views about participation in talk and their interpretations of classroom discussions. Based on observations and in-depth interviews, Clarke shows that for many students (especially those generally underserved by the education system), participation seems risky. In a troubling finding, she reports that many students remained silent (though not disruptive) for the entire observed six weeks of instruction in an urban high school. She raises questions about what nonparticipation means for student achievement, and discusses the need for students’ silence to be part of the research conversation about dialogic instruction.

In the final chapter of this set, Christine Pauli and Kurt Reusser (Chapter 14) discuss a large-scale Swiss-German research study that focused on finding relationships among characteristics of classroom talk and student performance. They found that both the quality of classroom talk and the quality of the presentation of mathematical content independently predicted student performance gains (see Applebee, Langer, Nystrand, & Gamoran, 2003, for similar studies in English language arts). They also provide evidence that teaching using academically productive talk was limited to students who were already high achieving. In effect, the kind of teaching that this volume suggests can produce intelligence and academic competence tended to be available only to students who already possessed that competence—a form of Matthew effect in which the rich get richer and the poor get poorer.

Section 3: Dialogue in the Digital Age

The chapters in Section 3 examine how academic discussion may shape (and be shaped by) the growing role of technology in education. Here, we include chapters that give us a taste of some of the variety in this booming new field of inquiry.

James Gee (Chapter 15) shows how computer-mediated games might provide a lens for interpreting classrooms that are trying to adopt dialogic methods. In the out-of-school, online interactions that Gee describes, participants shape and enforce common values concerning the content and processes they are engaged in. These shared views and skills, often absent in classrooms in which teachers begin to use dialogic practice, are essential to the game environment—and also, by implication, to the successful use of talk-based classroom pedagogies.

In formal education, students usually have little choice over what, how, when, and with whom they learn. In Chapter 16, Christa Asterhan reports on the use of computer mediation in classroom-located small-group discussions. Asterhan draws a distinction between an epistemic function that computers can perform to manage joint knowledge construction and a social function that manages who speaks and how students respond to one another. Drawing on findings from communication and learning sciences research, she predicts that what constitutes productive talk and productive teacher support in computer-mediated environments may prove to be quite different from their face-to-face counterparts.

In Chapter 17, Gerry Stahl shows how the careful design of theory-driven pedagogical software may create meaningful learning interactions between physically distributed learners. Stahl bases his argument on a series of studies with the Virtual Math Teams software, which school-aged children can use to engage in voluntary after-school mathematics activities. He suggests that the same basic prerequisites for productive learning interactions are present in online collaborations as in face-to-face collaborations: co-attention, shared understanding, and group cognition.

Allan Collins and Barbara White (Chapter 18) conclude this section with a reflective discussion on the general implications for education of the continuing migration of human communication to digital media. They suggest that digital communication technology does seem to have the potential for making dialogue central to learning beyond face-to-face classroom talk. They also warn about pitfalls, such as the danger of increasingly shallow communication. However, they argue that, somewhat paradoxically, digital communication opportunities can strengthen the central role that school and teachers play in socializing children to become reasoned speakers.

Section 4: Theoretical and Methodological Accounts of Learning and Development Through Dialogue

Section 4 raises two questions: (a) What are the social and cognitive processes through which structured academic dialogue builds intellective competence? and (b) How can we explain these effects? In the first subsection, scholars from different disciplinary backgrounds propose theories.

In Chapter 19, Anna Sfard takes up the first part of our challenge. She takes us on a conceptual journey through the arguments and claims made about the relationship between talk and learning, allowing us a kind of meta-vantage of the assumptions and ideas presented in this volume. She argues that thought is internalized dialogue, and self-dialogue is the means through which learning occurs. Participating in externalized dialogue, therefore, cannot magically make learning occur. The learner must also engage in an internal dialogue to make sense of talk.

James Greeno (Chapter 20) asks which of the processes that are engaged through dialogue support learning, retention, and possibly the transfer of knowledge in and across domains. As a psychologist drawing on sociolinguistics, he examines talk sequences and shows that they have an underlying cognitive structure. Sequences frame interaction, and thus they imply certain verbal responses. He argues that some sequences may be more productive than others, as they require students to engage in deeper cognitive processes.

In Chapter 21, Michelene Chi and Muhsin Menekse attempt to unpack the processes through which dialogue improves retention and understanding of academic content in peer interactions. They argue that gains from peer interaction depend on the extent to which partners engage in co-construction, a term that is often cited in the literature but left unspecified. An operational definition of co-construction is proposed that distinguishes it from other, less productive types of cognitive engagement. This allows for detailed and specific hypotheses to be formulated and (con-)tested empirically.

In Chapter 22, Kenneth Koedinger and Eliane Stampfer Wiese take on the challenge of explaining long-term improvement and far transfer effects, through the lens of cognitive science. They start with the argument that it is unlikely that a general faculty (such as intelligence) improves through classroom dialogue. Instead, they believe improvements happen in specific components of skill or competence that are imitated, practiced, and refined through repeated classroom dialogue opportunities. They also argue that any serious, scientific attempt to explore these processes in depth requires extensive collection of data that tracks growth over time and across a variety of settings and tasks.

New developments in methodologies and designs allow at-scale research on dialogic classroom learning and instruction. Technological developments in data harvesting and machine learning offer further opportunities for research at scale. However, automated analyses of large data sets of dialogue come at a price. The remaining three chapters in this section deal with the affordances and trade-offs between achieving an accurate understanding of the specific, on the one hand, and the generalities that are obtained from large data sets, on the other.

This tension is perhaps best formulated in the chapter by Carolyn Rosé and Alla Tovares (Chapter 23), who explore the conflict between the goals and methods of sociolinguistics, which aims to preserve complexity in the analysis of social interaction, and those of computational linguistics, which aims to find well-defined, stable patterns in data. Rosé (a computational linguist) and Tovares (a sociolinguist) juxtapose the two approaches and show how they may complement each other. Both authors advocate investigating the nature of the social processes and then letting these insights drive the search for meaningful patterns in data.

The quantitative study of the processes of dialogue involves multilayered, nested data. To study these data at scale, new statistical methods are needed. In Chapter 24, Ming Ming Chiu shows how Statistical Discourse Analysis enables researchers to identify social and cognitive processes that unfold during dialogue. Chiu’s method allows us to model the processes in talk and the ways in which they nest within larger social structures. His analysis also relates these processes to individual learning outcomes.

Richard Correnti, Mary Kay Stein, Margaret Smith, Jimmy Scherrer, Margaret McKeown, James Greeno, and Kevin Ashley (Chapter 25) argue that the systematic measurement of the quality of classroom discussions can also be used to increase the incidence of high-quality classroom discussions by offering professional development opportunities. For example, teacher-tailored visualizations of classroom discussion features can support teachers’ reflection on their practice. Thus, methodological advancements in measurement have the potential to become powerful tools for teacher learning, as well as providing a means to track teacher improvement over time.

Section 5: Scaling Dialogic Practice Through Teacher Development

The final set of chapters is about scaling practice as opposed to its measurement. Here we mean moving beyond individual superstar teachers to the widespread use of dialogic teaching in classrooms. We also mean moving beyond proof-of-concept studies and research methods rooted in detailed case studies (what could be called the anthropology of classroom talk), to larger samples of learners and teachers.

This section focuses on how we might make dialogic instruction a reality across contexts, for all populations of learners. All of these chapters acknowledge the complexity of orchestrating discussion when discussion by its nature is improvisational. How do we teach teachers to pay attention to student ideas and how they relate to instructional goals, and keep the conversation moving in fruitful directions that support whole-class subject matter learning?

In the first chapter of this set, Robyn Gillies (Chapter 26) shows that professional development interventions can have a positive impact on dialogue quality and on students’ subsequent learning outcomes. Her findings are particularly noteworthy because the intervention she studied was relatively short. Though the teachers were volunteers, they were not star teachers, as is often the case in descriptive studies of best practice.

By contrast, in Chapter 27 Sarah Michaels and Catherine O’Connor highlight the complexity and expertise necessary to orchestrate dialogic instruction effectively. Taking on the challenge of how we might scale dialogic instruction in all settings, with all students, the authors take us through the development of their research and training on Accountable Talk (Michaels, O’Connor, & Resnick, 2008). They suggest that professional development should support teachers in conceptualizing talk moves as tools, a repertoire of utterances that teachers can call on to address particular problems of practice in interaction.

Magdalene Lampert, Hala Ghousseini, and Heather Beasley offer a different, albeit complementary, perspective on professional development for preservice mathematics teachers. In Chapter 28, Lampert and colleagues ask how we might help novice teachers develop dialogic pedagogy. As they conceptualize it, professional training should position emerging teachers as sense makers about mathematics instruction in the same way that dialogic classroom instruction positions students as sense makers about the subject matter under discussion.

In Chapter 29, Mary Kay Stein, Randi Engle, Margaret Smith, and Elizabeth Hughes offer a model of teacher development of dialogic instructional practices in mathematics, based on what they refer to as the five practices for orchestrating productive mathematics discussions. In this chapter, they report on a case study of a cohort of graduate students engaged in learning how to notice and bring forward important student ideas.

In Chapter 30, Vivian Mihalakis, Anthony Petrosky, and Stephanie McConachie give us a bird’s-eye view of structured class discussions in English language arts, in which students interpret literature through a process of collaborative reasoning. The authors argue that such discussions must be carefully structured and coordinated. Planning lessons at this level provides a scaffold for teachers new to this way of teaching. At the same time, these lessons scaffold the development of students’ disciplinary habits of mind and provide multiple entry points for students to negotiate access to discussions.

In Chapter 31, Jonathan Osborne reviews a body of work on teacher development efforts that aimed for transformational change. He outlines the difficulties of working with teachers who may not be initially committed to the values of dialogic teaching and learning. Osborne alerts us to the complexity of the required change, which may involve overall practice, ideology, and beliefs about teaching. He warns us not to expect fast uptake when we look for robust changes on a large scale.

Lindsay Clare Matsumura and Helen Garnier meet the challenge of scale head-on. In Chapter 32, they report on an ambitious, district-wide professional development effort to embed dialogic pedagogy in reading instruction through a coaching program. Their data indicate that the quality of text discussions improved and student achievement rose significantly, including dramatic increases for English language learners. The authors also report on the systemic features that may have afforded some of these improvements, and others that may have constrained them.

Epilogue: Discussion Chapters

In Chapter 33, Robin Alexander offers both a commentary and a call to action. We find ourselves at a new crossroads, and Alexander poses critical questions for us to consider. This volume brings together the evidence—distilled, convincing, and compelling—yet there is still a gap between the evidence and practice. He points to the role of policy and the challenge of convincing decision makers that classroom dialogue is not idle chatter, but the means through which we can socialize intelligence.

Lauren Resnick reflects on how far scholars and practitioners have come in nearly half a century of studying talk as a means for learning, and considers how we might move forward from this point (Chapter 34). As this volume shows, the evidence indicates a tight interconnection between how we talk about things and how we come to know, learn, and teach. In other words, participation in discussion is a core act of learning. Resnick argues that we must acknowledge that fact—and see where it leads when we treat this way of learning as a basic human right.

References

Applebee, A. N., Langer, J. A., Nystrand, M., & Gamoran, A. (2003). Discussion-based approaches to developing understanding: Classroom instruction and student performance in middle and high school English. American Educational Research Journal, 40(3), 685–730.

Bakhtin, M. M. (1981). The dialogic imagination: Four essays. Austin: University of Texas Press.

Bourdieu, P. (1991). Language and symbolic power. Cambridge, MA: Harvard University Press.

Dewey, J. (1916). Democracy and education. New York: Free Press.

Doise, W., Mugny, G., & Perret-Clermont, A.-N. (1975). Social interaction and the development of cognitive operations. European Journal of Social Psychology, 5(3), 367–383.

Dweck, C. (2006). Mindset: The new psychology of success. New York: Random House.

Gee, J. (1989). The narrativization of experience in the oral style. Journal of Education, 171(1), 75–96.

Heath, S. B. (1983). Ways with words: Language, life and work in communities and classrooms. Cambridge, UK: Cambridge University Press.

Koedinger, K. R., Corbett, A. T., & Perfetti, C. (2012). The Knowledge-Learning-Instruction framework: Bridging the science-practice chasm to enhance robust student learning. Cognitive Science, 36(5), 757–798.

Ladson-Billings, G. (2009). The dreamkeepers: Successful teachers of African American children. San Francisco, CA: Jossey-Bass.

Mead, G. H. (1967). Mind, self, and society: From the standpoint of a social behaviorist (Vol. 1). Chicago: University of Chicago Press.

Michaels, S., O’Connor, C., & Resnick, L. B. (2008). Deliberative discourse idealized and realized: Accountable talk in the classroom and in civic life. Studies in Philosophy and Education, 27(4), 283–297.

Resnick, L. B., Levine, J., & Teasley, S. D. (1991). Perspectives on socially shared cognition. Washington, DC: American Psychological Association.

Resnick, L. B., Säljö, R., Pontecorvo, C., & Burge, B. (Eds.). (1997). Discourse, tools, and reasoning: Essays on situated cognition. Berlin, Germany: Springer Berlin–Heidelberg.

Reznitskaya, A., Kuo, L. J., Clark, A. M., Miller, B., Jadallah, M., Anderson, R. C., & Nguyen-Jahiel, K. (2009). Collaborative reasoning: A dialogic approach to group discussions. Cambridge Journal of Education, 39(1), 29–48.

Steele, C. M. (1997). A threat in the air: How stereotypes shape the intellectual identities and performance of women and African-Americans. American Psychologist, 52(6), 613–629.

Vygotsky, L. S. (1962). Thought and language (E. Hanfmann & G. Vakar, Eds., Trans.). Cambridge, MA: MIT Press.

Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes (M. Cole, V. John-Steiner, S. Scribner, & E. Souberman, Eds.). Cambridge, MA: Harvard University Press.

P R O L O G U E

The Study of Classroom Discourse: Early History and Current Developments

HUGH (BUD) MEHAN

University of California, San Diego

COURTNEY B. CAZDEN

Harvard University

In Learning Lessons (Mehan, 1979), we reported one of the earliest qualitative studies of classroom discourse. The study was the product of a unique collaboration between two university researchers: Courtney Cazden, who returned to a primary classroom for one year (1974–1975) as the teacher, and Hugh (Bud) Mehan, who observed and analyzed her teaching and wrote the book. Learning Lessons described the two basic structures of classroom lessons (Instruction-Reply-Evaluation and Extended Sequences), the turn-taking procedures that teachers deploy to maintain the flow of classroom discourse, and the skills that students acquire—often implicitly—in order to contribute as competent members of the classroom community.

In this chapter, we describe the origins and purposes of the Cazden-Mehan collaboration, the basic structure of classroom lessons we uncovered, and recent developments in the study of classroom discourse that have blossomed since then. This chapter was originally presented as an oral conversation at the 2011 conference on which this book is based, and in the first section, we retain some of the first-person informality, mostly in the plural. We occasionally use a separate I for either Cazden or Mehan as indicated.

How the Project Started

We met at the Social Science Research Council Conference on Language, Society, and the Child at the University of California, Berkeley, in the summer of 1968, coming there with different intellectual biographies.

Cazden: I started out as a public school teacher in a working-class neighborhood in Connecticut (1954–1961). My first- and second-grade students came from stable families—White, African American, Puerto Rican—whose fathers had good blue-collar jobs. Our elementary school functioned well, but other teachers and I wondered why so many of our students ended up in the lowest high school tracks.

Given the post-Sputnik urgency about school achievement at that time, public voices suggested that students’ language, especially Black dialect, might be part of the problem. Intrigued by such claims, I entered a doctoral program at the Harvard Graduate School of Education (HGSE) in 1961 and became excited by new developments in cognitive psychology and linguistics. I participated for several years in Roger Brown’s pioneering longitudinal study of the language development of three young children: two from graduate-student families and one whose parents had only high school education (Brown, 1973; Cazden, 1965).

By the early 1970s, now on the HGSE faculty, I wanted to see if my primary school teaching would be different, given my new understanding of language development. So I asked Bud Mehan if he could arrange a teaching assignment in a working-class elementary school, and if he would be interested in documenting my teaching.

Mehan: At the time of Courtney’s request, I was directing the teacher education program at the University of California, San Diego, and was able to arrange for her to teach in a primary grade classroom in the urban core of San Diego for the 1974–1975 academic year. Her combined first-, second-, and third-grade classroom was composed of Latino and African American students referred to her by other teachers at the school.

I approached the prospect of documenting Courtney’s classroom instruction from an amalgamation of academic orientations. Trained as a sociologist with a specialization in ethnomethodology (Mehan & Wood, 1975), I focused on social practices such as special education groupings, ability groupings, and tracking systems. My constructivist orientation in sociology was reinforced by concurrent developments in linguistics, anthropology, and psychology. Chomsky’s (1968) theory of linguistic competence dominated the field of linguistics at the time, but it had been challenged by sociolinguists and psycholinguists¹ for neglecting the social origins and functions of language. Their studies of communicative competence were echoed by anthropologists such as Frake (1964) and Goodenough (1964), who framed culture in terms of participation or membership in a society—what one has to know, believe, and especially do, in order to operate in a manner that is acceptable to the members of a community or a society. I wanted to extend that logic to the classroom by asking, What do students have to do in order to be seen as competent members of the classroom community? I hoped to answer that question by examining Courtney’s interactions with the students in her classroom.

Thus, our interests coalesced in this collaborative project, and we were both energized by a shared concern for educational equity. During that period, the gap in academic achievement among Blacks and Whites was often explained by a deficit theory. In the biologically based version of deficit thinking (e.g., Jensen, 1969), Black youth were portrayed as cognitively inferior to their White counterparts because of racial differences. In the culturally based version (e.g., Deutsch, Katz, & Jensen, 1968), the inferior speech and thought of low-income Black and Latino youth were blamed on their families’ cultural practices. We hoped that a study of classroom discourse might provide an opportunity to examine, and perhaps challenge, such deficit thinking (cf. Cazden, 1976, 2001). This study was among the first to use videotape as a tool to study the social organization of complete events in educational settings, such as classroom lessons, testing sessions, and counseling sessions.² We videotaped the 1st hour of school activities every day for the 1st week of school and then 1 hour a day approximately every 3rd week until April. This procedure produced a corpus of materials with nine instances of a similar event.

The Structure of Classroom Lessons

Teacher-led lessons are the most prevalent form of events that occur regularly in classrooms. (Others include individual instructional time, sharing time, small collaborative groups, presentations of learning, and laboratory experiments.) Classroom lessons have unique organizational features that distinguish them from other classroom events and from ordinary conversations that occur outside of classrooms. One of their most salient features is their three-part sequence. A teacher’s initiation act induces a student’s reply, which in turn invokes a teacher’s evaluation. For example, early in the school year, when children were still learning one another’s names, Cazden taught a lesson that produced the following exchange.³

Segment 1. Name Cards Lesson

Evaluations are more likely to occur in classroom interactions than in other situations in everyday life, because in everyday conversation, speakers routinely ask questions to obtain information they do not possess (information-seeking questions). The presence of an evaluation is one of the features that distinguish classroom conversations from those that occur in everyday situations.

The three-part Initiation-Reply-Evaluation (IRE) structure exists in large part because of the language game (Wittgenstein, 1953) organized by the teacher. In the recitation language game, the teacher begins a sequence looking for a simple, preferred answer; the student responds (with a short answer); and the teacher evaluates the answer. The third part in the IRE sequence is crucial for controlling the flow of interaction. This is where the correctness or incorrectness of a student’s reply is established. If confusion or unintelligibility is not cleared up, it is assumed that the student’s reply was acceptable.

In the recitation game, teachers often ask for information they already know—known-in-formation questions (Mehan, 1979; Shuy & Griffin, 1978; Sinclair & Coulthard, 1975). Not all teacher-led initiation acts test students’ previously acquired knowledge, however. In another class of initiation acts, teachers seek new information, ask students for their opinions, or ask for interpretations. When information-seeking questions are introduced into classroom lessons, much longer, or extended sequences of interaction often transpire. Because the teacher seeks information from students he or she does not know, lessons can proceed in unexpected directions, as the following exchange between Cazden and her students illustrates.

Segment 2. Birthplaces Lesson

When the topic shifts to Prenda’s mother, a more extended sequence unfolds:

Cross-Cultural Variations and Educational Recommendations

Researchers who study language use in homes and schools have suggested that classroom lessons may be compatible with the discourse patterns in middle-income families but incompatible with the discourse patterns of certain minority group families. This discontinuity, in turn, may contribute to lower achievement in U.S. schools among students from African American, Hawaiian American, Native American, and Latino backgrounds.

Researchers have recorded variations on the default condition of short IRE sequences (Cazden & Mehan, 1989) in a variety of settings, often with students and teachers from cultural groupings that differ from the mainstream.⁴ Au and Jordan (1981) showed that a Polynesian teacher used discourse techniques that differ from the basic IRE sequence when teaching Native Hawaiian students in the Kamehameha Early Education Program (KEEP). In the introductory part of a lesson, designed to elicit students’ knowledge about frogs before students read about them, the exchanges in Segment 3 occurred after the teacher asked, What would you do with a frog if you captured it?⁵

Segment 3. Freddie Found a Frog

This teacher, like Cazden in the birthplaces lesson, is asking information-seeking questions. She does not know beforehand what her students propose to do with a captured frog. Note that she explicitly asks for a range of answers (see Turn 3.2) and that more than one student answers her questions before she evaluates their responses. Even when she identified Cyndy as the responder, she received a range of answers—from benign and helpful (feed it, put it in a pond, put it in a bucket) to more mischievous (poke the legs). Researchers associated with KEEP point out that these discourse features, especially those that enable students to assemble answers jointly, are consistent with discourse patterns in Native Hawaiian communities (Au, 1980; Au & Jordan, 1981; Tharp & Gallimore, 1988).

Erickson and Mohatt (1982) revealed similar patterns in a classroom of Algonkian Indian children taught by an Algonkian Indian teacher on the Odawa reserve in Canada. That teacher asked known-information questions, but her manner of allocating turns for students to talk in the Initiation frame was different than in conventional classrooms: she allowed students to select themselves to speak next. This turn allocation technique does not obligate students to stand out as individuals; instead, they can choose to participate in unison as members of groups when answering questions. Erickson and Mohatt point out that her evaluation moves were different also. She did not evaluate students’ replies after each turn; she dispensed praise in public, criticism in private. The researchers applauded the use of these techniques for being more compatible with the cultural style of the Native students than the more individualized IRE sequences of traditional classrooms (cf. Philips, 1983).

These authors also observed a non-Indian Canadian teacher in the same school. When he first started teaching, his style was similar to the default IRE condition in all respects. The longer he taught on the Odawa reserve, the more his teaching practices shifted in the direction of his students’ culture. In effect, his students informally socialized him to use more group-based instruction, facilitate more voluntary contributions, and keep evaluations private as they worked together over time. His students implicitly taught the non-Indian Canadian teacher culturally compatible discourse strategies.

Some researchers have tutored teachers explicitly in how to use culturally compatible discourse strategies. In her study of Trackton, Heath (1983) reported that Trackton teachers used instructional strategies with low-income students that were not prevalent in the students’ homes. Whereas Trackton teachers would use known information questions in IRE formats, Trackton parents would use statements or imperatives. And when parents did ask questions of Trackton children, their questions called for nonspecific comparisons or analogies as answers. Heath concluded that the language used in Trackton homes did not prepare children to cope with the major characteristics of the language used in classrooms: utterances that were interrogative in form but directive in pragmatic function, including known information questions and questions that asked for information from books.

Heath worked with Trackton teachers on ways to appropriate the community’s style of interacting with children. After reviewing tapes of parent-child interactions with researchers, teachers began social studies lessons with questions that asked for personal experiences and analogic responses, such as What’s happening there? Have you ever been there? What’s this like? These questions were similar to the questions parents asked their children at home. Their use during the beginning of lessons generated active responses from previously passive and nonverbal Trackton students. Once the teachers increased the students’ participation in lessons using home-based questioning styles, they were able to use more school-based questioning styles.

This line of research has led to several educational recommendations. The basic idea is to use students’ home knowledge and cultural practices as a resource in classroom instruction. Teachers are encouraged to build upon the strengths of home language and culture. This may involve initially modifying the participation structures of the classroom to be more compatible with the participation structures of the home and then moving on to socialize children to mainstream classroom talk.

The work reviewed so far in this section shows a concern to enable certain underrepresented minority group students to participate more actively in classroom lessons. The studies we review in the next section are concerned with ways to encourage students to gain control over the complexities of academic discourse. The work of Teresa McCarty, Carol Lee, and their colleagues represents a transition from a concern for participation to a concern with cognitive development within studies of classroom discourse.

For decades, Native American students have been portrayed in the literature as quiet, passive, and nonresponsive. They have been said to learn by observing and doing, not through listening and saying (More, 1987; Tharp, 1989). Often in the name of cultural compatibility, educators have emphasized nonverbal means of instruction and cue-response scripted drills as ways to reach passive Native American students. McCarty et al. (1991) stated that these erstwhile attempts have had an unfortunate side effect. Native American students are taught neither to engage in reasoning with evidence nor to use inquiry methods.

Working with the Navajo-staffed Native American Materials Development Center, staff members of the Rough Rock Indian reservation school implemented an inquiry-based bilingual social studies program. Students were asked to scour their community and return with lists of things the community needed. Students were then asked to group like items and justify their choices. Eventually they reached a consensus, identifying things needed and things they’d like to have. That consensus led to the lesson generalization: Rough Rock is a community because people work together to meet their needs and solve mutual problems (McCarty et al., 1991, p. 57). The lessons in Navajo humanities suggest that Navajo students will indeed respond enthusiastically to inquiry-based questioning that asks them to provide evidence for their observations.

In a similar manner, Lee (1995) described how six teachers in two urban high schools helped their predominantly African American students increase their skills of interpreting rich literary texts by introducing and building upon material drawn from cultural routines and rituals. The teachers first helped their students analyze the rules for achieving irony in signifying rituals. Next they invited students to read and analyze contemporary literature, such as Toni Morrison’s (1970) The Bluest Eye and Alice Walker’s (1982) The Color Purple. During these instructional units, teachers regularly asked questions that required students to make inferences, interpret figurative language, and draw upon evidence from disparate parts of the text as well as personal experiences of the social world the texts embodied. A comparison of students’ scores on pre- and posttests of text interpretation reveals students’ shift to more sophisticated interpretations of complex inferential questions.

From Recitation to Reasoning

Classroom lessons composed primarily of IRE sequences have been criticized for being overly teacher centered, for converging on one correct response, and for demanding factual information or yes-or-no responses from students (Edelsky, 2006; Macbeth, 2003; Wells & Arauz, 2006). Partially in response to such criticisms, we have witnessed a shift in focus within the study and practice of classroom discourse. The studies revealing cultural variations on the recitation script we reviewed in the previous section advocated for increasing the participation of minority students in classroom lessons in the name of educational equity. If minority students could be encouraged to take more turns at talk in classroom lessons, they could be better prepared to contribute more actively in the full dimensions of school life. More recent studies we review in this section are concerned with prompting minority students