layout-8/l-2003 1/7/04 12:29 PM Page 19 Conversation As a Model of Instructional Interaction Joan Van Bramer, State University of New York at Albany ABSTRACT The role of social context and the nature of human interaction provide rich resources for the study of learning and human cognition. In order to under- stand these elements more fully, it is important to consider the language in use within these contexts. The early intervention Reading Recovery is grounded in the belief that the conversation between the teacher and the student during the reading and writing of texts is the stimulus for learning. In this study, discourse analysis in the form of conversation analysis was used to analyze transcripts of the introduction to the new book from two Reading Recovery lessons. An intertextual analysis was also conducted on one of the transcripts to evaluate the conversational genre versus the traditional classroom discourse pattern of initi- ate, respond, evaluate (IRE). Results showed that one transcript sample was a clear example of a conversational discourse style with the teacher and student actively co-constructing meaning, while the other was an example of the more traditional IRE style with the student maintaining a passive stance to learning. Implications and limitations of the study are discussed. Literacy Teaching and Learning Volume 8, Number 1 pages 19–46 layout-8/l-2003 1/7/04 12:29 PM Page 20 Literacy Teaching and Learning Volume 8, Number 1 Teaching is essentially a matter of facilitating learning, and where that learning depends on communication between the teacher and the learner, the same principles apply as in any successful conversation. The aim must be the collaborative construction of meaning, with negotiation to ensure that meanings are mutually understood. (Wells, 1986, p. 101) The role of discourse in learning, particularly in educational settings, cannot be understated. Analyzing discourse helps us understand how knowledge is created within human social interactions and, as Jaworski and Coupland (1999) sug- gest, deconstruct how particular texts come to be structured as they do, with all their social and political implications. From this perspective, language use is viewed as a form of social practice (Fairclough, 1999). Understanding discourse in this way may help educators provide the instructional conversations most beneficial to all students. The purpose of this paper is to, through discourse analysis, consider the principles of conversation within an educational framework. Specifically, I will examine the interactions within a one-to-one tutoring setting to determine whether or not the principles of conversation are being applied and to describe how meaning is being co-constructed. THEORETICAL FRAME My understanding of conversation as instruction is informed by Jaworski and Coupland’s (1999) research regarding conversation analysis and research that describes what is meant by conversation within educational frameworks. In the following two sections, I provide background on these perspectives. Conversation Analysis Jaworski and Coupland (1999) describe several key features of conversation analysis. These features include openings and closings, adjacency pairs, topic management and topic shift, conversational repairs, showing agreement and disagreement, and mechanisms of turn taking which they say is probably the most central. Within this social practice, there are well-established scripts that speakers follow with clearly defined roles. In adjacency pairs, for example, two utterances are linked, with the first actually defining how the second utterance is constructed and with the anticipation that the second will fill that slot as expected. Such pairs might include question-answer, greeting-greeting, or offer- acceptance/refusal. In these examples the first utterance leads to a specific type of utterance in response. 20 layout-8/l-2003 1/7/04 12:29 PM Page 21 Conversation As a Model of Instructional Interaction Van Bramer Conversation, then, is highly structured in that each participant builds upon the previous utterance and assumes that the other party will do the same. There is an understood responsibility that each participant will check for understanding across the conversation. This shared understanding leads the speaker to repair a misunderstanding if one occurs and the listener to request that repair if needed. Jowarski and Coupland go on to describe conversation analysis as residing in a sociological approach to language and communication known as eth- nomethodology. They define ethnomethodology as “a means of studying the link between what social actors ‘do’ in interactions and what they ‘know’ about interaction” (1999, p. 19). The goal of conversation analysis is to discover and describe the organization of social interaction and to understand better how this organization reinforces the structures of social organization and institu- tions. Thus, they consider conversation as “talk in interaction” (1999, p. 19). Talk, as described by these authors, is a matter of sharing meaning-making pro- cedures that they view as the starting point for conversation analysis. Schegloff and Sacks (1999) describe the organization of speaker turns, not- ing that an organizational problem of every conversation is figuring out how the exchange will end. The solution that speakers use, they noted, was to use closings, which refers to saying something that both participants mutually understand to mark the end of the exchange. Another feature of conversation that Schegloff and Sacks explore is the con- cept of what they term topic shading, or shifting the topic of conversation. This shift is managed without specifically ending the topic but by fitting in differ- ently focused, but related, talk to some last utterance. It is important to observe how shifts in topic are achieved and, in terms of social contexts, to consider who appears to control this topic management. The work of Bakhtin (1999) emphasizes the active role of the listener and the co-construction of meaning by the speaker and listener. He clarifies the role of the listener as a person who understands the utterance and takes an active and responsive stance to it. It is through this response that the listener becomes the speaker. From this perspective, dialogue can be messy, depending on the social context and the previously held meaning by the listener or speaker (Peyton-Young, 2000). According to Bakhtin, speakers choose a particular speech genre when speaking, and within the genre, there are stable and typical forms of construct- ing the message. Genres are diverse and the speaker may be unaware of them, yet genres organize speech in a manner similar to grammatical forms. Finally, in terms of conversation theory, Schiffrin (1999) discusses the use of conversational markers as a form of information management. For example, “oh” in a conversation can convey a variety of meanings within a conversational exchange. It can be used to signify a cognitive realignment, to represent the 21 layout-8/l-2003 1/7/04 12:29 PM Page 22 Literacy Teaching and Learning Volume 8, Number 1 speaker’s shift to the role of listener, to signal consensus or understanding, or it can signal a disjunction and surprise at encountering a different point of view or knowledge state. Conversational Interactions Within Educational Frameworks Over the past two decades there has been an increased focus upon the sociocul- tural aspect of learning and cognition. As Rogoff (1990) relates: “A broader view of cognition and context requires that task characteristics and cognitive performance be considered in the light of the goal of the activity and its inter- personal and sociocultural context” (p. 6). The child is no longer viewed as a learner in isolation but as an active participant in the construction of knowl- edge within social contexts. As a result of this focus, the importance of language and social interaction for cognitive development has been recognized. Rogoff (1990) considers children as apprentices in thinking, active in their efforts to learn from observing and participating with peers and more skilled members of society. They use cultural tools to develop problem-solving strate- gies within the context of sociocultural activity. Through guided participation, children and others work collaboratively to build bridges between the novice’s present understandings and new ones. The “more knowledgeable other” arranges the child’s participation in activities, shifting responsibility toward the child as capability increases. The dialogue and nonverbal actions within the interaction are critical to this concept of guided participation. The conversa- tions during such interactions are viewed as leading to continued cognitive development (Nelson, 1996; Rogoff, 1990; Vygotsky, 1978). Moll (1994) discusses the importance of an interactive, meaning-based approach for the teaching of literacy. He describes participatory approaches that highlight children as active learners using and applying literacy as a tool for thinking and communication. These approaches emphasize the role of the teacher in facilitating the child’s taking over the learning activity. Again, from this perspective, the discourse practice related to the instructional interaction becomes critical for success. As such, conversation becomes a tool for mutually constructed meaning. Clay (1998) identified several features that conversation and teaching inter- actions have in common. Both require cooperation between the speaker and lis- tener, a speaker who has a goal of making ideas clear, and listeners who, as con- tributors to the interactions, bring their own information to bear on the con- versation. In both conversation and instructional interactions, speakers and lis- teners check for understanding and reformulate the message as needed. Just like partners in a successful conversation, both teacher and student need to be active participants in the cooperative construction of meaning. As Clay says, When we speak we do not assume that our listeners cannot think; we 22 layout-8/l-2003 1/7/04 12:29 PM Page 23 Conversation As a Model of Instructional Interaction Van Bramer expect them to bring knowledge to bear on what we are talking about. Yet probably the most common error made by adults about the learn- ing of young children is that we can bypass what the child is thinking and just push new knowledge into the child.…However, any learning situation is like a conversation, for it requires the learner to bring what he or she already knows to bear on the new problem being explored.… If we become observers of our own conversations—noting when they go well, when they get into difficulties, how we negotiate over our dif- ficulties, and when and why communication fails—this may help us understand a little better how children learn. (p. 15) Clay views the teacher as bearing the responsibility for creating opportuni- ties for rich conversations within instructional interactions and suggests that by using models of conversation and communication as a guide, teachers can improve on their ability to use the understandings that children already have. It is through building upon these understandings that children are able to make meaningful connections to incorporate new knowledge. In their paper on teachers’ instructional actions, Roehler and Duffy (1991) discuss the shift in emphasis in instruction from the teaching of rote memoriza- tion of facts and skills to a higher level of understanding through actively con- structing knowledge. They quote Resnick (1981) as saying: Today’s assumptions about the nature of learning and thinking are interactionist. We assume that learning occurs as a result of mental constructions of the learner. These constructions respond to informa- tion and stimuli in the environment, but they do not copy or mirror them. This means that instruction must be designed not to put knowl- edge into learner’s heads, but to put learners in positions that allow them to construct well-structured knowledge. (Resnick, 1981 cited in Roehler & Duffy, 1991, p. 862) The conversation within the interaction allows for this construction of knowledge. The teacher’s scaffolding of information within this conversation assists the student to use prior knowledge and integrate new learning. As Roehler and Duffy (1991) state, “scaffolding normally occurs in dialogues of structured interactional sequences in the midst of instruction” (p. 872). Wilkinson and Silliman (2000) describe the range of teacher-student dia- logues as a continuum of interrogation sequences to instructional conversations, and they note that it is through the supportive scaffolding of instructional con- versations that students become flexible and reflective self-regulated learners. They quote Tharp (1994) saying, “The critical form of assisting learners is through dialogue, through the questioning and sharing of ideas and knowledge that happens in 23 layout-8/l-2003 1/7/04 12:29 PM Page 24 Literacy Teaching and Learning Volume 8, Number 1 instructional conversations.…To truly teach, one must converse, to converse is to teach.” (cited in Wilkinson and Silliman, 2000, p. 343) Finally, Cazden (1988) describes classroom discourse as “a kind of applied linguistics—-the study of situated language use in one social setting” (p. 3). Her research revealed a traditional style of teacher and student interaction, consist- ing of a three-part sequence. First, the teacher initiates by calling upon a spe- cific student; second, that child provides a narrative response; and third, the teacher comments on the response providing a form of evaluation. This initiate, respond, evaluate (IRE) sequence is the most dominant form of speech interac- tion at all grade levels, and it is one in which the teacher controls who speaks and when and does the majority of the talking with little interaction between students. As an alternative to this pattern of discourse, Cazden suggests another way to think about educational talk that, with small shifts away from the traditional IRE discourse pattern, can have a significant cognitive and social impact. One important shift that she emphasizes is a change from the typical recitation for- mat to a model of discourse closer to real discussion, where the instructional talk becomes more like informal conversation. In order for this change to occur, she presents three discourse features that must be altered. The first is what she calls speakers’ rights.This requires a shift to more self-selection by students in terms of the management of turn taking. The second feature is the teacher’s role, in which the teacher moves away from con- stant questioning to other forms of input such as declarative statements, reflec- tions, and invitations for students to elaborate. The final feature is speech style, in which the teacher learns to use a more exploratory form of talk that is less planned and more responsive to student input. This review of research suggests that teachers can develop instructional interactions that meet the diverse needs of all students by becoming more thoughtful about instructional discourse features and adjusting talk to take into account what students understand. In the next section, I review research that deals specifically with conversation within Reading Recovery, the context for this study. Conversational Interactions Within Reading Recovery Reading Recovery is an early literacy intervention designed for children experi- encing extreme difficulty learning to read. Interactions during reading and writ- ing receive a particular focus during Reading Recovery teachers’ professional development sessions in which they study effective teaching. In fact, the heart of the instruction is said to be “conversations between the teacher and child while involved in reading or writing” (Klein, Kelly, & Pinnell, 1997, p.166). Effective Reading Recovery teachers provide opportunities to negotiate 24 layout-8/l-2003 1/7/04 12:29 PM Page 25 Conversation As a Model of Instructional Interaction Van Bramer meaning through talk (Lyons, Pinnell, & DeFord, 1993). They include the fol- lowing quotation from Clay (1985): “Sensitive and systematic observation of young children’s reading and writing behaviors provides teachers with feedback which can shape their next teaching moves. Teaching can then be likened to a conversa- tion in which you listen to the speaker carefully before you reply (p.6).” (p. 58) Through sensitive observation and skillful conversational interactions, the the- ory goes, even the most confused children can become successful. In summary, our understanding of the role of conversation in educational frameworks is informed by research in anthropology, sociology, and linguis- tics—disciplines that include the study of human interaction. Within the con- text of social interaction, discourse allows for the co-construction of meaning. The use of discourse analysis, then, can be a powerful tool in understanding the construction of knowledge in educational settings. Next I describe the research methodology of this study and how the data were analyzed. The preceding review of conversation analysis as one form of critical discourse analysis and conversational interaction within educational frameworks will serve as a basis for the analysis. METHODOLOGY The study was situated within two elementary schools, both of which used Reading Recovery. One school is located in a small town of approximately 17,000 and the other in a small rural community. Both schools are in the northeastern region of the United States and serve a majority population of Caucasian students. Approximately 46% of the children in one school and 34% in the other school qualify for a free or reduced-price lunch. Data were collected within the context of Reading Recovery lessons. The lessons take place daily for a period of 30 minutes. The focus of the study was on the concluding 10 minutes of instruction when the teacher introduces the new book to the student. Participants Two teachers were selected for the study along with two of their students. Both teachers spend approximately half of their day teaching Reading Recovery stu- dents and the other half of their day working with small groups of Title I reme- dial reading students. Both teachers are veteran educators and have at least 3 years of experience as Reading Recovery teachers. Each teacher was selected because of her history of positive outcomes with students and her ability to be reflective about her practice. 25 layout-8/l-2003 1/7/04 12:29 PM Page 26 Literacy Teaching and Learning Volume 8, Number 1 The two students in the study were selected randomly from each teacher’s caseload of Reading Recovery students. They were selected for Reading Recovery in the fall of their first-grade year because of their status as being among the population of students experiencing the most difficulty with learn- ing to read and write. Data Sources Participant observation, field notes, audiotaping, and transcription served as the research tools. The researcher observed and audiotaped the teacher’s book intro- duction and the conversation between the teacher and the child during the introduction. Field notes were used mainly as a tool for comparison since the audiotapes and transcripts provided much more detailed information. As an observer, the researcher sat behind the teacher and child in an attempt to be as unobtrusive as possible. Data Analysis Data that were gathered through participant observation, field notes, audio- tapes, and transcripts were analyzed in two phases using discourse analysis. The first phase included the conversation analysis of transcripts of the book intro- duction and the teacher-child verbal interaction during the introduction. The second phase involved an analysis of intertextuality for one of the transcripts. During the first phase, the transcript for each teacher and her student was analyzed for characteristics typical of genuine conversation. Features such as mechanisms for turn taking, openings and closings, adjacency pairs, topic man- agement and topic shift, conversational repairs, and showing agreement and disagreement were noted for each of the two transcripts. I explored additional evidence of conversation by searching for instances of these three things: partic- ipants checking for understanding, showing signs of cooperation, and building upon previous utterances. The second phase of the analysis considered the nature of intertexuality of one of the transcripts. The goal of the analysis was to understand how two dif- ferent discourse genres appeared within one lesson excerpt and to consider how that might have affected the instructional conversation. The transcript was assessed for evidence of the traditional initiate, respond and evaluate (IRE) instructional genre (Cazden, 1988) and a less typical conversational instruc- tional model. Specific consideration was given to speakers’ rights, the teacher’s role, and the teacher’s speech style—elements suggested by Cazden as having the potential for significant positive cognitive and social impact on students. 26 layout-8/l-2003 1/7/04 12:29 PM Page 27 Conversation As a Model of Instructional Interaction Van Bramer FINDINGS Characteristics of Conversation in New Book Orientations The goal of the teacher’s book introduction is to make the book accessible to the child. The new book is carefully selected based upon the child’s current capabilities so that there are just a few new things to learn, and it is carefully introduced to provide a framework of meaning to scaffold the successful read- ing. As stated by Clay and Cazden (1990), “What may seem like casual conver- sational exchanges between teacher and pupil are based on deliberate teaching decisions for a particular child” (p. 210). Therefore, the conversation during the book introduction is critical for the joint construction of meaning that will allow the child to read the book suc- cessfully. Teacher modeling of how to actively construct meaning from text dur- ing the introduction will allow the child to take over this task while becoming a constructive rather than passive learner. Appendix A explains the transcription conventions, and Appendixes B and C contain the transcripts that were analyzed. In the first set of findings, the analysis refers to specific lines in each of the two transcripts for evidence or lack of evidence of the characteristics of conversation to be discussed. The second set of findings demonstrates the elements of intertextuality in terms of the tradi- tional instructional discourse of IRE versus the less institutionalized discourse of everyday conversation. Conversation Analysis of Transcript 1 As mentioned earlier, conversation analysis centers upon certain characteristic features of conversation. Specific features are demonstrated from transcript excerpts below. Bracketed conversation indicates the teacher and child are speaking at the same time, while = indicates contiguous utterances and ( ) indi- cates talk that is unheard or is in doubt. The transcript is arranged to show when the overlapping occurred. The numbers correspond to the lines in the transcript included in Appendix B. In Transcript 1, Lines 1–3, the teacher (T) has a clear opening which makes a connection with an earlier book read by the child (C) : 1 T: Now, you remember we’ve been reading an ABC book, child’s name, about the zoo?= 2 C: = Yeah. 3 T: And all the animals in the zoo? This sample also shows an adjacency pair that could be considered 27 layout-8/l-2003 1/7/04 12:29 PM Page 28 Literacy Teaching and Learning Volume 8, Number 1 summons-answer, followed by a number of such pairs in Lines 5–8, which begin to get the child involved in the book: 5 T: Well, this is another book called We Went to the Zoo.What are some of the animals you think you would see at the zoo? 6 C: Elephants! 7 T: The elephant. You find elephants at the zoo, don’t you? 8 C: Yeah. There are other examples of adjacency pairs. For example, in Lines 81 and 82, the pair provides an example of compliment-compliment, and Lines 27–28 and 46–47 show what might be classified as offer-acceptance pairs: 81 C: [I’m the man.] 82 T: [You’re the man!] 27 T: …Should we look and see what kind of animals? 28 C: Sure. 46 T: …Let’s see what else 47 C: [( )= (The child overlapped but couldn’t be understood and was contiguous with the next line.) There are clear examples of turn taking. Early in the sample the teacher seemed to control most of the talk, but as the child became more engaged, the talk became more like everyday conversation, with the child beginning to initi- ate more talk and even overlapping the teacher talk in places such as in Lines 20, 22, and 41: 19 T: [They have stripes too. Well that’s going to make it 20 C: [They have um in their back. 21 T: Oh, well you know a lot about – cheetahs. 22 C: [Animals.] 40 T: Ahh. W- 41 C: [It doesn’t bite. The overlaps and contributions suggest the child’s engagement with the topic and desire to make personal connections. There appears to be a sharing of topic management between teacher and child. The teacher was receptive to the child’s contributions and responded positively showing that these contributions were valued. This is a strong example of cooperation to create meaning because it is characteristic of everyday conversation. Conversation also involves cooperation for the co-construction of meaning. Evidence of this cooperation is seen in how the participants build upon each other’s utterances. For example, in Line 9 the teacher makes a statement, and in Line 10 the child builds upon the teacher’s utterance by linking back with the pronoun it.The teacher then (Line 11) builds upon the child’s comment and 28