Abstract
Despite warrants for classroom discussion, research consistently finds that discussions in K-12 classrooms remain rare. Our research investigates whether and in what ways practice-based teacher learning opportunities focused on discussion facilitation influence opportunities for student talk. Grounded in data from a job-embedded professional development program for fourth- to eighth-grade literacy teachers, we analyzed videos of teachers co-planning and co-facilitating discussions with students. Findings indicate that, across 1 year, during co-facilitated classroom discussions, teacher talk decreased, while student talk increased. In a parallel finding, our analysis of co-planning sessions revealed that the ways teachers planned for discussions also changed. Teachers went from engaging only in what we call proactive pedagogical reasoning to balancing that with what we have come to call responsive pedagogical reasoning.
In dialogic discussions, students engage in collective sensemaking as they construct interpretive meaning with texts (Lee, 2007; Levine, 2014). Scholars of democratic education frequently tout dialogic discussions as a tool for advancing educational justice and the democratization of knowledge (Hess, 2009; Levinson, 2012; Parker & Hess, 2001). Along with being tied to the democratic purposes of schooling, research has linked classroom discussion to a wide variety of desirable student outcomes including reading comprehension, student engagement, and conceptual understanding (Applebee et al., 2003; Engle & Conant, 2002; Kazemi & Stipek, 2001; Nystrand, 1997). In English language arts (ELA), the benefits of discussion have been particularly well researched. Most recently, a 2021 study found that the number of conversational turns in a lesson predicted statistically significant improvements in young children’s learning outcomes (Neuman et al., 2021) and a 2019 study of 639 ninth-grade ELA students found that the more students talked in class, the better they performed on a reading literacy test. This link applied regardless of socioeconomic background or gender (Sedova et al., 2019). These recent findings echo earlier studies that have found that discussion-based approaches to literacy learning improve students’ ability to engage independently in literacy tasks (Applebee et al., 2003).
Given the preponderance of empirical and theoretical warrants for classroom discussion, both scholars and practitioners remain frustrated by the rarity of classroom discussions in schools (Mehta & Fine, 2015; Saye & The Social Studies Inquiry Research Collaborative, 2013; Wiggins, 2015). In one study, Nystrand et al. (2003) found that the ninth-grade students they studied engaged in only 15 seconds of classroom discussion per day. While Nystrand and colleagues’ findings were particularly stark, research spanning decades has found that discourse-rich, student-centered instruction is rare (Cazden, 2001; Cohen, 1990; Goodlad, 1984; Kane & Staiger, 2012; Mehta & Fine, 2015; Saye & The Social Studies Inquiry Research Collaborative, 2013; Sizer, 1984; Wiggins, 2015). The persistence of these disheartening findings raises a simple, but puzzling question: how can teachers be supported to deepen their practice when it comes to facilitating discussions in ELA classrooms?
Simple on its surface, this question is, in fact, exceedingly complex. It forces us to face the reality that cultures of schooling continue to value silence and obedience from students, qualities and behaviors that are at odds with high-quality discussions. As historians of education have repeatedly illustrated, these cultural values in K-12 schools have persisted despite decades of reform efforts oriented toward student-centered instructional paradigms that value discussion and dialogue (Cohen, 1988; Cuban, 1986). While long-held cultural value-systems are difficult to shift across all professions, cultures of schooling and teaching are particularly difficult to transform because new teachers enter the profession having experienced almost two decades of enculturation into professional norms during their experience as students (Lortie, 1975). Therefore, the stubborn absence of classroom discussion in K-12 schools is not a simple problem to unpack. However, it is one that scholars and practitioners have been working hard to unravel, especially in recent years.
One community of scholars that has been trying to investigate how teachers might be supported to do the hard work of shifting cultural norms in teaching by embracing student-centered classroom discussion are scholars studying practice-based teacher education (PBTE) and professional development (PBPD) (Alston et al., 2018; Farrow et al., 2022; Ghousseini et al., 2021; Grossman et al., 2019; Kavanagh et al., 2022). Practice-based approaches to teacher learning anchor teachers’ learning experiences in specific teaching practices, like discussion facilitation (Ball & Forzani, 2009; Grossman, 2018; Grossman et al., 2009, 2021; McDonald et al., 2013). Anchored in particular practices, teacher learners engaged in PBPD experience cycles of enactment and investigation that include the representation of a practice, preparation to enact the practice, the enactment of the practice, and an opportunity to analyze their enactment (McDonald et al., 2013).
Scholars arguing for practice-based designs for teacher learning have typically grounded their arguments in sociocultural theories of teacher learning (Cole, 1996; Leont’ev, 1981). Such theories place particular importance on the ways that learning processes are culturally and historically situated (Scribner & Cole, 1996). By beginning with such assumptions, PBTE scholars acknowledge the pervasive influence that harmful cultures of schooling have historically had on teachers’ practices. They suggest that without firmly rooting teacher learning experiences in the activity of the classroom, it is likely that long-standing cultures of schooling will wield more influence over teachers’ practice than any professional development program ever could (Grossman et al., 1999). This is because the frameworks that guide teachers’ thinking and activity are developed through their practice—through the problem-solving activities they enact in culturally and historically situated communities of practice (Wertsch, 1981). By grounding teachers’ opportunities to learn firmly inside of their professional practice, PBTE scholars hope to acknowledge the pervasive nature of cultures of schooling while also, hopefully, shifting those cultures (Grossman & McDonald, 2008).
While the body of literature on PBTE is growing, debates about the approach have recently emerged. One growing critique of a practice-based approach to supporting teachers is that by working with teachers on predetermined and prespecified practices, instruction may become more scripted and less responsive to students and to context (Kennedy, 2016). For those who ascribe to this critique, focusing on practices walks a line dangerously close to prescriptive teaching. As Daniels and Varghese (2020) note, “practice-based teacher education often reflects the belief that teaching and learning can and should look the same and produce relatively uniform results for all students and teachers” (p. 5). While many PBTE scholars would disagree with the argument that specification of practice is the same as standardization of practice (Danielson et al., 2018; Grossman et al., 2009; Hauser & Kavanagh, 2019; Kavanagh, 2017, 2021; Metz et al., 2020), for critics of PBTE focusing on a specified teaching practice precludes the kind of responsive approach to instruction that ambitious forms of discussion facilitation require. Taking these critiques seriously, we designed an inquiry that would help us understand whether the practice specification work that PBTE requires led teachers in one PBPD to become more standardized in their instruction and less responsive to their students. Our research focused on the following questions:
As teachers collaboratively planned discussions during PBTE experiences, to what extent and in what ways did they plan to be responsive to contexts and to students?
When teachers did plan to be responsive to students and to contexts in their lesson plans, how did these plans play out when they attempted to enact core practices in classrooms with children?
Theoretical and Conceptual Perspectives
Dialogic Discussion
Our interest in classroom discussion is grounded in a preponderance of research illustrating text-based discussions are pivotally important to student learning in the ELA classroom (Almasi & McKeown, 1996; Applebee et al., 2003; Kucan & Palincsar, 2013). The importance of discourse to students’ literacy learning has been well documented in studies of how students are supported to understand and analyze literature (Evans, 2002; Kucan & Beck, 2003; Nystrand & Gamoran, 1991; Pennell, 2014) as well as studies of how students develop critical consciousness around identity and social issues (Campano et al., 2013; Gambrell et al., 2011; Handsfield & Crumpler, 2013; Rangu et al., 2021). In a study spanning 80 schools, Applebee and Langer (2009) found that dialogic classroom discussion was significantly related to performance tasks requiring students to adopt interpretive stances in literature. This and other studies have illustrated that, through discussion, students often move from scanning text for “the point” to exploring and negotiating possible meanings—a more sophisticated and more humanizing approach to meaning-making with literary text (Langer, 2010).
Along with empirical warrants, the power of discussion in the literacy classroom is grounded in theoretical warrants from Vygotskian sociocultural theories of learning to Bakhtin’s theories of the dialogic. For Vygotsky, social interaction is the groundwork atop which all meaning-making occurs. The work of understanding and comprehending is an inherently social event, not an event that takes place within an individual. In fact, Vygotsky (1978) even uses the word dialogue when discussing the social work of negotiating meaning and developing understanding. Picking up the threads of Vygotsky’s forays into dissecting the relationship between dialogue and understanding, Bahktin pushes these ideas even further by emphasizing how the intersection of multiple voices is always already at the center of the work of meaning-making. In fact, for Bakhtin (1981), there is no escaping dialogue, even when alone. He is also clear, however, that some contexts do a poor job of setting the stage for true dialogue, tending, instead, toward monologic conversations which resist the unpredictable and dynamic nature of dialogue (Barnes & Todd, 1995). To push back on monologic classroom contexts (Bakhtin, 1981), many educators have called for a “dialogic classroom,” one in which there are more egalitarian social relations between teachers and students when it comes to the work of meaning-making, where students interact freely with each others’ ideas, and where understanding is an unfolding process negotiated by all participants rather than a pre-scripted path laid out by a teachers’ lesson plan (Billings & Fitzgerald, 2002; Nystrand & Gamoran, 1997; Wells, 1999).
Responsive teaching
Promoting true dialogic discussion requires teachers to understand their work in ways that often counter cultures of schooling, which are often grounded in assumptions that the job of the teacher is monologic: to transmit knowledge to a passive learner (Jackson, 1986). In contrast, dialogism in classrooms requires teachers to be responsive to what their students contribute to the unfolding classroom activity. Dialogic discussion requires teachers to see their students as doing much more than simply receiving knowledge. Instead, students are negotiating meaning with teachers and peers through their participation in culturally and historically situated discourses and within particular communities of practice (Gee, 1989; Lave & Wenger, 1991; Rogoff, 2003). Seeing learners and learning through this sociocultural lens, the work of teaching expands, requiring teachers to be in relationship with students’ emergent ideas about content, to respond authentically to those ideas, rather than simply replace them with new ideas. This perspective is often referred to as responsive teaching (Hammer et al., 2012; Kavanagh, Metz, et al., 2020). A responsive approach to teaching contrasts with instructional approaches in which all or most aspects of activity are planned in advance, the goal of which is to begin lessons with gapless visions of all that will be said and done (Duffy, 2000). During responsive teaching, while many important decisions are made in advance of instruction, most teacher actions are made in response to what students contribute. The teacher’s role, therefore, is not to provide knowledge to students, but rather to create opportunities for students to grapple with complex texts, problems, and tasks and then to engage in strategic uptake of students’ ideas in ways that push thinking forward (Nystrand et al., 2003; Nystrand & Gamoran, 1991). Prioritizing responsiveness in teaching requires believing that students enter classrooms as competent sensemakers with existing emergent understandings that are worthy of engagement (Lampert et al., 2013).
Pedagogical Reasoning
The role of responsive teachers in dialogic discussions is significantly less dominant than the role of a teacher would be in a classroom where their role is to dispense knowledge. However, while the role of a teacher may appear less central during responsive instruction in dialogic classrooms, it is not less complex (Kavanagh, 2016, 2018). This is in part because the dynamic and unpredictable nature of dialogic interaction means that teachers cannot fully pre-plan their engagements with students. The inherently responsive nature of teachers’ roles in dialogic discussions means that their in-the-moment sensemaking—what Shulman (1987) calls pedagogical reasoning—becomes exceedingly complex, requiring teachers to call on knowledge to be responsive to the diverse needs of learners in a particular moment and context and in service of particular goals (Loughran et al., 2016). As students’ ideas unfold in dialogic discussions, teachers are faced with a series of what Lampert (1985) has called pedagogical dilemmas: moments when teachers are faced with a choice to which there is no one clear answer. These dilemmas require teachers to engage in reasoned arguments with themselves (e.g., who to call on, when to stay silent). Managing these dilemmas, Lampert argues, is the constant work of teaching.
Mansfield (2019) recently investigated teachers’ reasoning processes as they engage these dilemmas. She argues that when teachers encounter dilemmas, their state of pedagogical equilibrium is disrupted causing a chain of events, which, together, make up the work of pedagogical reasoning. In her work with Loughran (2018), she offers a framework that unpacks the chain of events that make up pedagogical reasoning. We found this framework useful in our analysis of teachers’ pedagogical reasoning during dialogic discussions. Mansfield and Loughran’s (2018) framework identifies three stages of pedagogical reasoning: the situation, the reaction, and the consequence. Situations are moments when teachers encounter dilemmas that challenge their sense of equilibrium. Reactions to those situations involve “an intellectualization of the dilemma (Dewey, 1933) which can lead to problem setting (Schön, 1983)” (Mansfield & Loughran, 2018, p. 2). Consequences of that intellectualization are the modes of action that bring teachers back to the pedagogical equilibrium they seek.
In our previous research, we have visualized this process of pedagogical reasoning as a cycle of dilemmas, judgment, decisions, and moves, in which every move creates a new pedagogical dilemma, which starts the cycle again (Kavanagh, Conrad, et al., 2020). For this study, we draw both on our own previous work and on the work of Mansfield & Loughran to understand teachers’ reasoning processes as they engage in dialogic discussion. Our visual representation of this cycle can be found in Figure 1 and our operationalization of these concepts in analysis is unpacked in our “Method” section.
Teacher reasoning during dialogic discussion.
Perspectives, in Sum
There are rich theoretical and empirical warrants for expanding and deepening classroom discussions, particularly those that do not shy away from the unpredictable and dynamic nature of dialogue. However, to promote egalitarian classrooms where understanding is treated as an unfolding process rather than a pre-scripted path, teachers must be responsive to the ideas that students bring. For teachers, responsive teaching requires a significant amount of pedagogical reasoning as they make sense of students’ emergent ideas and then use these ideas as fuel to power classroom activity. We embarked on this investigation to understand whether and in what ways teachers in a year-long practice-based teacher learning experience were able to work on being responsive to students and context as they focused on developing their practice as discussion facilitators. And further, whether and in what ways their practice as discussion facilitators changed across the year of their participation.
Context for the Study
This study was conducted during the first year of a 2-year school-university partnership between a research university and a public K-8 school in a large city in the United States, which, for the purposes of this article, we will call The Center School. The Center School is a rapidly changing school. Over the past 10 years, the school has more than doubled in size and has gone from a school where almost 100% of students qualified for free and reduced-price lunch to one where the district identifies only 80% of students as “economically disadvantaged.” This change is due to the surrounding neighborhood’s rapid gentrification. The school serves many immigrant families, with 22.3% of students classified by the district as “English learners.” As of 2016, there were 17 languages represented in the student body. Along with serving a linguistically diverse student body, the school also represents a diversity of racial identities: 17% of families at the school identify their child as Black/African American, 30% identify their child as Hispanic/Latino, 34% identify their child as White, 11% identify their child as Asian, and 8% identify their child as multiracial.
Our project engaged The Center School’s literacy teachers in Grades 4 through 8 with the authors of this article (who we will often refer to as “the university team”) in learning labs, a job-embedded, practice-based teacher learning experience. During learning labs, the university team and the participating teachers spent an entire school day together (on lab days the participating teachers’ classrooms were supported by substitutes). Lab days followed a predictable pattern: First, we investigated the purpose of dialogic discussion, then we collectively developed a lesson plan centered around dialogic discussion, next we went into a classroom to co-facilitate our planned discussion with students, and finally we left the classroom and debriefed the experience together. During some labs, our group would facilitate discussion in two classrooms instead of one, changing the lesson plan slightly in between classroom visits. This learning sequence—(a) introduce a practice (in this case, dialogic discussion), (b) plan to enact the practice, (c) enact the practice with students, (d) analyze enactment—was drawn from the learning cycle (McDonald et al., 2013). For labs, the learning cycle was adapted to allow all stages of the cycle to occur in one day at one school site as a collective experience in which teachers teach together rather than in their separate classroom contexts. While the learning labs model was originally developed by math educators to support math teacher learning (Gibbons et al., 2017), in our work, we adapted the model to support literacy teacher learning and focused labs specifically on dialogic discussions grounded in short literary texts, such as poems or vignettes.
The authors of this article were all members of the learning lab facilitation team. When data for this project were being collected, in the 2018–2019 academic year, the first two authors were faculty in schools of education and the third author was a doctoral student. All identifying as White, female, teacher educators with backgrounds in K-12 teaching, our interest in studying teacher learning during school-based learning labs arose out of our previous research. In other investigations, we had studied models of PBTE and PBPD that took place outside of K-12 school contexts (Grossman et al., 2018; Kavanagh & Danielson, 2020; Kavanagh et al., 2019; Kavanagh & Rainey, 2017; McDonald et al., 2013; Reisman et al., 2019) and we embarked on this study interested in seeing how the pedagogical approaches to teacher education that we had been analyzing in university settings would work in K-12 spaces. Our interest in the relationship between practice-based pedagogies and teacher responsiveness arose as we began to think deeply about the critiques of PBTE that were published around the time of data collection (Daniels & Varghese, 2020; Kennedy, 2016; Philip et al., 2019).
Method
Participants
Teachers
All fourth- through eighth-grade teachers who engaged in text-based discussions at The Center School were invited to enroll in the 2-year experience and all teachers who accepted this invitation were enrolled in the study. In addition, all preservice teachers (PSTs) who were placed in the classrooms of enrolled teachers were invited to participate and each of these three PSTs accepted the invitation. In total, our project enrolled eight participants, which included two fourth-grade teachers, three fifth- and sixth-grade literacy teachers, one eighth-grade social studies teacher, and three PSTs. Three of the teachers identify as men (all of whom identify as white) and six of the teachers identify as women. Of the six women, three identify as White, one as African American, and two as Asian American. The teachers’ years of experience ranged from preservice to 24 years, with the majority of the group having taught for more than 10 years.
Data Sources
Our primary data for this analysis includes videos and transcripts of the planning enactment portions of the first five of six learning labs from the first year. The sixth learning lab was excluded from this analysis because it had a different structure. During labs, video was captured by positioning a stationary camera in the room. Videos captured teachers and the university team in conversations during collective planning (totaling over 15 hr of video, see Table 1).
Lab Planning Videos.
Video also captured teachers, the university team, and students during the instructional enactment portion of labs. Each lab had one or two enactments for a total of nine enactments in Labs 1 to 5, totaling over 7 hr of video-taped instruction (see Table 2).
Length of Lab Enactment Videos.
For this analysis, all lab videos were professionally transcribed, totaling over 30 hr of video.
Data Analysis
Curious about how students’ classroom talk changed over the course of the year in the classrooms where teachers were engaged in collaborative instruction during learning labs, we closely analyzed student talk happening in each of the nine lab enactment videos across the five learning labs (during four out of the five learning labs teachers visited two different classrooms, resulting in nine enactment videos across five learning labs). Using deductive coding, we coded each video for conversational turns. A conversational turn begins when a participant (teacher or student) speaks and includes a response from another participant (teacher or student). A conversational turn was coded each time the speaker changed after a response. This process allowed us to generate data about how frequently students had opportunities to speak in the classroom.
Interested not only in changes to classroom talk, but also in how teachers’ pedagogical reasoning during planning may have been related to changes in students’ opportunities to talk, we looked across the lab transcripts asking: As teachers collaboratively planned discussions during PBTE experiences, to what extent and in what ways did they plan to be responsive to contexts and to students? To code, we began by identifying all pedagogical dilemmas: moments where two (or more) alternative pedagogical choices were considered by teachers as they collaboratively planned.
For each dilemma, we used codes derived from our theoretical frame (see Table 3). In particular, we operationalized Mansfield and Loughran’s (2018) conceptualization of pedagogical reasoning to code the moment the dilemma emerged (the situation), the discussion and intellectualization of the dilemma (the reaction), and the mode of action or choice made (the consequence) (see Table 1). We used Mansfield and Loughran’s (2018) framework to write memos for each dilemma in which we identified the situation, the reaction, and the consequence.
Coding Scheme and Examples.
Finally, given that our research question was particularly focused on understanding how teachers planned to be responsive to students, we sampled only those dilemmas that arose during the labs planning segment. Examining the situation, reaction, and consequence of each of these dilemmas a final set of coding categories emerged: proactive pedagogical reasoning and responsive pedagogical reasoning (see Table 1). We coded moments when teachers resolved a pedagogical dilemma prior to enacting the lesson (e.g., deciding which text to read) as proactive pedagogical reasoning because teachers were making pedagogical decisions prior to entering the classroom and we coded moments when teachers intentionally left pedagogical dilemmas unresolved as responsive pedagogical reasoning because (as we discuss in the findings), when dilemmas were considered and left unresolved, teachers often acknowledged that those dilemmas needed to be resolved with input from students. The emergence of the proactive and responsive pedagogical reasoning codes and our analysis of patterns in how these codes were applied over time is what led to the findings that we report in the following section.
Findings
The larger project in which this study is embedded focused on supporting a group of preservice and inservice literacy teachers as they developed their capacity to be responsive to students during dialogic, text-based discussions. For this particular investigation, we wanted to understand (a) whether and in what ways the nature of students’ talk changed over the course of the project and (b) whether and in what ways teachers’ planning for dialogic discussions changed over the course of the project. What we found was twofold. First, we examined trends in student versus teacher conversational turns during all of the enactment portion of labs. What we found was that, over the course of the year, the number of teacher conversational turns in discussions steadily decreased while the number of student conversational turns in discussions steadily increased. As a result, the classrooms went from places where teachers were doing the majority of the speaking to classrooms where students were doing the majority of the speaking.
To understand what may have contributed to this favorable change, we turned to examining trends in how participants planned for these discussions and there we also found a significant change over time, this time the change was in what teachers did when they encountered pedagogical dilemmas while planning. Early in the year, teachers focused their planning on resolving a series of dilemmas that needed to be figured out prior to entering the classroom. We call this proactive pedagogical reasoning and it includes definitively answering such questions as: Which text should we read? What should we write on the anchor chart? What norms should we set? Later in the year, when teachers encountered pedagogical dilemmas, they much more frequently intentionally left their dilemmas unresolved, what we call responsive pedagogical reasoning. In the first learning lab of the year, teachers resolved 100% of the pedagogical dilemmas they encountered during planning and by the final lab the split between resolved and unresolved dilemmas was almost half and half. In the following sections, we unpack these findings before offering our thoughts about their significance for practitioners and for future research on teaching and teacher education. We turn first to our findings about changes in student talk during discussions.
How Discussions Changed: From Majority Teacher Talk to Majority Student Talk
To understand how the nature of classroom talk changed over the year, we coded each conversational turn in all our classroom enactments during learning labs. In total, 1,674 conversational turns were identified across nine classroom discussions. The average number of conversational turns per classroom enactment was 190 and ranged from 107 to 318 (see Table 4).
Turns of Student and Teacher Talk by Enactment.
We coded each conversational turn as either “teacher” or “student,” which indicated whether the speaker was a teacher or a student. Next, we examined our coded data for chronological patterns and found that, over time, student talk increased and teacher talk decreased (see Figure 2).
Talk over time.
Figure 2 highlights the changing pattern in talk as we see the increasing trend line for student talk crossing over the decreasing trend line for teacher talk, as student talk increased teacher talk decreased. An examination of this graph illustrates that changes in student and teacher talk were not steady and linear. However, the trends represented in Figure 2 show a clear pattern that, over time, the percentage of student conversational turns increased in comparison to the percentage of teacher conversational turns. In each of the first four videos, more of the conversational turns were coded as teacher than as student, whereas in three out of five of the final videos more of the conversational turns were coded as student than as teacher. In addition, the first and the last lab are almost exact inverses of each other. In the first lab, out of a total of 203 conversational turns, 124 were teacher turns and 79 were student turns. In the final lab, out of a total of 197 conversational turns, 68 were teacher turns and 129 were student turns.
In addition to these observed quantitative changes, we also observed qualitative changes to student and teacher talk throughout labs. In the early labs, teachers spoke in between nearly every student, and due to the collaborative nature of instruction during labs, it was not uncommon for more than one teacher to speak between each student. In these early labs, we observed teachers doing most of the cognitive work of the discussion, such as asking the questions, probing student thinking, connecting student ideas to each other, and generally pushing the discussion forward. Observing student talk from these early labs, we see short responses—often amounting to no more than one word answers or short phrases, students speaking to the teacher rather than to each other, and little evidence of students connecting their ideas to those of other students.
These qualities of talk are illustrated in an excerpt from our first lab enactment, which took place in an eighth-grade literacy classroom and engaged students in a discussion of the text Code Breakers, a chapter from the book Speak by Laurie Halse Anderson. The text—written from the perspective of a middle school student—captures the frustration experienced by students trying to “break the code” of literary texts for their teachers. The following excerpt from this lab shows teachers dominating the discussion by posing all the questions, directing student talk, and doing most of the cognitive work. At the same time, student comments are short, disconnected from each other, and fail to richly engage the text or other students’ ideas. The excerpt begins after a student had shared that they identified with a specific character from the text: Rachel Rochelle. The teacher elevated this comment for discussion: Teacher 1: Please raise your hand if you felt like you connected with Rachel Rochelle [a character in the text]? Similar to [Student 1]. Please raise your hand. Okay. Who here connected with the narrator? [Who] was like, “Oh, reading is kind of fun?” Who here connected with Hair Woman [the teacher in the text]? No one? I heard some people! Yeah, here? Would you, [Student 1], mind telling us why you connected with Hair Woman? Please listen! Student 2: Because she’s a reasonable person. Teacher 1: Can you speak up? Student 2: Because she’s a reasonable person. Teacher 1: Okay, so she says she connects with Hair Woman because she’s a reasonable person. Anybody have any ideas about that? Comments? Teacher 2: How about Hair Woman? . . . Does Hair Woman sound familiar? Who’s Hair Woman? Student 3: She’s a teacher. Teacher 2: Okay, she’s a teacher, right? . . .The way she’s speaking to the class, does it sound familiar? How come? Student 3: Because.
Notice in the excerpt that the teachers are doing the majority of the cognitive work in the discussion. In addition to posing all the questions, the teachers add their own elaboration to students’ short and disconnected comments. Students’ comments are brief and fall short of engaging each others’ ideas or the text. These qualities are not unique to this excerpt and are emblematic of the degree to which the teachers, not the students, carried the discussions in our early labs.
As labs progressed, we saw qualitative changes in talk that aligned with the collaborative work we had been engaged in. In later labs, we observe students engaging in most of the talk with teachers interjecting sparingly in response to what students have said rather than to guide the conversation in a particular way. When teachers did engage, they engaged as other participants in the discussion rather than the one leading the talk. Teachers spoke to ask students to elaborate on their thinking (e.g., “So say more,” “What do you mean?”) or a few times to ask them to speak louder so everyone has access to their thoughts (e.g., “Can you repeat what you said . . . I thought it was really fascinating but it was hard to hear back here.”). Examining student talk from these later labs, we see more elaborate, lengthier responses. Students were responding to and building on one another’s ideas, using the talk moves that had been introduced to them in labs (e.g., “I agree,” “Back to what ___ said”), and asking each other questions about shared comments. One example of this type of student talk comes from an enactment in a fifth-grade classroom where students were engaged in a discussion around the text Eleven by Sandra Cisneros. The text, written from the perspective of a girl on her 11th birthday, is about her feelings when a teacher insists a sweater in the lost and found is hers. During this discussion, students continually share, connect, and build on each other’s ideas without teacher interjection. The longest excerpt of student talk includes eight conversational turns. During one of these, a student connected her idea to a comment made by another student.
I agree with you. So on [lines] 17 or 15, I underlined it because I don’t think that it’s right for a teacher to judge a student because they’re a geek or the way they look. And because they believe I guess the other girl, I guess you could call her smarter, but maybe that’s not true. I don’t think that it’s great for a teacher to force a kid to put on something and not hear the kid out because the girl was shy to say that it’s hers, but it’s really not.
In this excerpt, we see that the student is connecting to another student’s idea, using the “agree” talk move, and engaging in rich talk that includes evidence from the text.
Buoyed by finding this favorable trend in conversational turns and an increase in rich student talk over the course of the year, we were interested to know whether we could find any parallel changes in how teachers planned for discussions. This interest led us to analyze our data from the collaborative planning sessions for these discussions. We report our findings from our data analysis about planning for discussions below.
How Planning for Discussions Changed: An Increase in “Responsive Reasoning”
The overarching project in which this study is embedded was focused on supporting a group of preservice and inservice teachers as they developed their capacity to be responsive to students during dialogic, text-based discussions. Perhaps it is unsurprising, then, that our next research question focused us on understanding the extent to which and the ways in which teachers planned to be responsive to context and to students as they planned for text-based discussions. In an analysis of how teachers planned for discussions, we found that as student talk increased and teacher talk decreased, the nature of teachers’ pedagogical reasoning during collaborative planning went through a parallel change. Early in the year, when teachers encountered pedagogical dilemmas while planning (e.g., which text should we select, which question should we ask, what norms should we set, etc.), they always resolved the dilemma before entering the classroom. Later in the year, by contrast, teachers frequently waited to resolve pedagogical dilemmas that arose during planning until their choice could be responsive to students and to the context. Through our analysis, we came to call resolved dilemmas “proactive reasoning” and unresolved dilemmas “responsive reasoning.” In the following sections, we first illustrate the change over time in teachers’ engagement in each type of reasoning during collaborative planning and then we unpack each type of reasoning using qualitative data to illustrate the difference.
Changes over time in teachers’ pedagogical reasoning during planning
Over the course of our collaborative work, teachers engaged in less proactive and more responsive reasoning in response to planning dilemmas; Figure 3 illustrates this finding. In our first lab, 100% of dilemmas engaged proactive pedagogical reasoning. This differs from the fifth lab where only 60% of dilemmas engaged proactive pedagogical reasoning and 40% were responsive pedagogical reasoning. The trend line of proactive reasoning decreasing and responsive reasoning increasing indicates that as teachers engaged in learning labs, they began to wait to resolve some pedagogical dilemmas until they could make decisions in response to students and context in-the-moment.
Reasoning type by lab.
Proactive pedagogical reasoning
Proactive pedagogical reasoning occurs when teachers resolve a pedagogical dilemma prior to enacting the lesson. In each lab, there were elements of the lesson that were always determined before entering the classroom. For example, in almost every lab teachers would engage in proactive pedagogical reasoning around text selection and the opening discussion question. Two examples of proactive pedagogical reasoning are provided below as illustrations of common dilemmas teachers discussed that were resolved prior to enacting the classroom with students.
The first example is a dilemma we saw arise across all labs, where teachers engaged in proactive pedagogical reasoning. A moment where the text had been selected but teachers were discussing how it would be read with students.
Teacher 1: So the first question, I think, is do we want to read the text as a whole without stopping first and then go back [and discuss portions]? Or do we want to read and stop and read and stop? Teacher 2: I would read it as a whole. Teacher 1: It’s pretty short. Teacher 3: Yeah. The first time through. Teacher 2: Yes. For one it’s pretty short, but also I feel like in ways towards the end, she [the author] kind of contradicts the beginning. It would be helpful, maybe, to read it straight through before we start [discussing] instead of start and stop and start and stop.
At this moment, teachers proactively made a decision about how the text would be read. As lessons typically started with reading the text, it was important for teachers to know what they would be reading and how it would be read. Along with text and reading style selection, teachers often discussed any resources they would want to prepare and have in the classroom during the lesson and made a decision prior to entering the space. The next example illustrates a dilemma that arose around anchor charts.
Teacher 1: Do we need any particular anchor charts? Teacher 2: I definitely think the talk moves. Teacher 3: I think talk moves. Teacher 4: Do we want to use the term “waiting” or “thinking” [on the anchor chart]? Teacher 2: I would call it maybe thinking time. Teacher 1: Yeah.
In this dilemma, one teacher noted it would be helpful to have a chart with the talk moves, referring to Chapin and colleagues’ (2009) five productive talk moves: waiting, revoicing, repeating, reasoning, and adding on. The talk moves had been introduced to students but there was not an anchor chart to reference in this particular classroom. Other teachers in the group have talk move anchor charts up in their rooms and have found the resource helpful for students to reference during discussions. The group had previously discussed that wait time for a teacher is time for students to think. Therefore, a few teachers had been using the terms thinking or think time with students instead of waiting. The group agreed to create an anchor chart to take into the classroom and to use the term thinking instead of waiting. In this dilemma, teachers decided to create the anchor chart ahead of time, being proactive.
While it may be common for teachers to engage in proactive reasoning around text and resource selection, as illustrated in these first two examples, we found that teachers also engaged in proactive reasoning in early labs in response to dilemmas that more overtly welcomed teacher responsiveness to student ideas, such as what questions to ask students if students are reluctant to speak. To illustrate such a moment, we look to the first learning lab. As previously described, the enactment for this lab occurred in an eighth-grade literacy class and involved students’ discussion of an excerpt from a chapter titled “Codebreaking” from the text Speak by Laurie Halse Anderson. In preparation for this lesson, the teachers did not plan to be responsive to students’ ideas. Instead, the teachers’ lesson plan consisted of seven different discussion question options that they could ask if the discussion stalled. Rather than planning to listen to student ideas and determine discussion questions in response to student thinking, their lesson plan was designed to provide options that could be summoned with little thought to push the discussion in a particular direction. When it came time to teach the lesson, the lab participants used the lesson plan in exactly this way. After struggling to generate student talk with the question “Who do you identify with [in the text] and why?,” one teacher hoped that a different question would stimulate conversation and referred to the list of discussion questions that they group had generated prior to instruction: Teacher 1: I think we need a [new] option. Like we need Option #2 or something. [referring to predetermined questions that the teachers had brainstormed prior to the lesson. Option #2 was “Why does the author end with the last line?”] Teacher 2: We can do [options] #2 and #5. Start with #2 and then go from #2 into #5 because they [the questions] tie into each other.
In this moment, the teachers did not consider student thinking during instruction when they encountered the decision of determining their next discussion question. This example of proactive reasoning during planning resulted in teachers asking questions that they were curious about, rather than considering what students were curious about and using that knowledge to determine the next discussion topic.
Each of these dilemmas were resolved with a specific action determined prior to entering the classroom. Early in the year, every time a pedagogical dilemma emerged, teachers engaged in this type of reasoning in response. Later in the year, teachers often engaged in proactive pedagogical reasoning, but it wasn’t the only type of reasoning they engaged in. They continued to make decisions about which text to read and what to write on their anchor charts and which question should launch discussions. However, in addition, they also engaged in what we call responsive pedagogical reasoning, which we unpack in the following section.
Responsive pedagogical reasoning
Responsive pedagogical reasoning occurs when pedagogical dilemmas arise during planning and teachers intellectualize those dilemmas by engaging with them, but ultimately leave them unresolved until decisions can be made in response to students and context in-the-moment. During the first lab, teachers did not engage in any responsive reasoning as they planned. Rather, every time they encountered a pedagogical dilemma, they made a clear decision that included a specific action that would be taken, resolving the dilemma. As the year progressed, this changed (see Figure 3). As student talk increased, teachers found the ideas children presented to classmates could be highlighted for discussion rather than just focusing on the questions teachers had entered the room with. Thus, planning shifted as they anticipated these moments arising. As teachers saw the benefits of responding to contexts and students in-the-moment, they became more comfortable resolving dilemmas without a particular move but rather the understanding that the decision needed to be made in-the-moment with students. We provide two examples below to illustrate moments of responsive pedagogical reasoning.
The first example arose after teachers had been discussing, earlier that morning, the lively and unpredictable nature of classroom discussions. Teachers reflected on past discussions they had led with students during learning labs in which students had surprised and impressed them with their interpretations of text and had left teachers unsure, in-the-moment, of where to go next. While planning to lead a discussion with students later that day, two teachers reflected on the conversation they had had as a group of teachers: Teacher 1: This morning we talked a lot about student responses that we didn’t anticipate or that we think are totally on the mark. We talked a lot about that. What do we want to experiment with [when we’re leading the discussion with students later today] in relation to the teacher action or inaction in relation to that moment that’s likely to occur when we are unsure of where to go? Teacher 2: I don’t know, but I think having a huddle. Most likely a moment will come up and even if we’ve said, like, “Let’s try and see where it goes,” I feel like if I was teaching at that moment, I’d have some unease. But I wonder about doing a huddle where we can huddle together and be like, “Okay, I don’t know.”
In this excerpt, a pedagogical dilemma emerges: What should we do when a student shares an idea that we didn’t anticipate and we feel unsure about what to do next? Rather than resolving it by making a clear decision about exactly what to do in that moment, Teacher 2 suggests that the teachers acknowledge when they are unsure during instruction and enlist support in making an in-the-moment decision at that time. Rather than being proactive, Teacher 2 suggests that the group wait to make a decision until they can be responsive to the specificity of the moment when it emerges. This decision to wait to see what arises before deciding a course of action is intentional.
This example highlights the group’s understanding that we cannot pre-plan all interactions with students during discussions given the dynamic and surprising nature of student talk. To authentically respond, judgments must occur in-the-moment. The group recognizes that there will be comments and ideas shared by students we cannot anticipate. Rather than deciding on a particular move to make when this occurs, teachers plan to respond in-the-moment.
The example above comes from our fourth learning lab. By the time it took place, the group had engaged in multiple labs and classroom discussions and had created a community of learners that felt comfortable co-planning and teaching. The above example differed from the early examples of responsive reasoning that began to emerge in the second and third labs. To illustrate how responsive pedagogical reasoning first transpired and developed over time, our second example is the first coded instance of responsive pedagogical reasoning which occured in our second learning lab. We include this instance to show how responsive reasoning changed over time as teachers collaboratively planned and engaged in text-based discussions with students.
As illustrated in the example above from the first learning lab, it had been common in early labs for the teachers to ask a question and for the students to respond with silence. As a group, we began referring to this silence as “crickets.” When planning for the second lab, the dilemma of what to do when we ask a question and no one responds surfaced.
Teacher 1: Let’s be really intentional about crickets [moments of silence]. Let’s imagine we ask the question, any one of these questions, and we get crickets. Teacher 2: I wonder if we might experiment with instead of asking a follow up question, doing something else like . . . everyone jot down one answer that you might have, and then share it with your partner. Because I worry a little bit that when we ask a question and there’s crickets, and then we ask another question, it starts to feel a bit like, “Oh God, I’ve got to get the right answer,” and then kids shut down even more. Teacher 3: I say to students, when they’re pair sharing, if I heard something really good, and I know it was a student who wasn’t going to participate, “I heard you tell so and so of this, do you mind if I share your answer?” Or, “do you want to share your answer?’If the kid tells me no, I’ll say, “do you mind if I share your answer?” Teacher 4: We have a few potential moves if the question results in silence.
This example illustrates an anticipated dilemma based on previous enactments and teaching experiences. The group predicts a moment will arise where students will be silent but are unable to anticipate what will precede that moment. We recognize the decision must be made in collaboration with the students given their input is necessary for action. Therefore, rather than coming to a consensus about a particular move to make in-the-moment, the group acknowledges there are multiple options that must be decided in-the-moment based on the context of the classroom and how students have been participating. Rather than making a specific decision about what to do if there is silence, the teachers surface a few possible moves they could employ in-the-moment with students.
This first instance of responsive pedagogical reasoning illustrates that as teachers became aware that there were moments they could not anticipate they initially discussed a few potential moves they could make, creating a tool box of moves to draw on in unpredictable moments. As time progressed, we saw a shift in this type of reasoning. Instead of discussing a few possible moves that might be made, teachers decide to make decisions collaboratively with themselves and the contributions of students in-the-moment. Overtime, teachers seemed to understand just how unpredictable and dynamic discussions may be and that decisions must be made in-the-moment from an infinite number of moves based on the specific context rather than a set number of options discussed ahead of time.
As teachers’ reasoning shifted to be more responsive over time, we saw a similar shift in talk, with student talk increasing. To provide an example of what may have contributed to an increase in talk, we share an excerpt from Lab 4 when a teacher uses student ideas to move the discussion forward.
Teacher 1: I just heard two really interesting thoughts from two students and I just wonder if it makes sense to state those again for the class to respond to? . . . I’m wondering if the two of you can say what you just said again? So . . . we’ve heard multiple ideas of what we think is going on here, but there were I think two really powerful thoughts just shared about what the poet or the author is trying to get at in this poem. I’m wondering if the two of you can each say what you said again? They’re not the same. So it’d be interesting to hear . . . your reactions to that. Like, “I think this is what’s going on.” . . .Or maybe you think something entirely different is what’s happening right now.
This example contrasts with what happened in the first lab, when the teachers selected discussion questions from a proactively planned list without considering student contributions. By highlighting comments from two students and asking them to share their thinking, the teacher engaged in responsive reasoning, rather than proactive reasoning. Instead of relying on the teachers’ proactively determined questions, student thinking was highlighted and discussed.
Our findings reveal a relationship between an increase in teacher responsive reasoning and an increase in student talk. While it is possible that the teachers’ increasing engagement in responsive reasoning across labs created the space for more student voices to be heard, our study design and data set do not allow us to make such causal claims. Nonetheless, these data provide us with information to begin to understand how an intentional plan to be responsive in-the-moment to students’ ideas may create more space for children to share their thinking and for their ideas to influence the direction of the discussion.
Discussion
We began this research by studying classroom talk in the context of a year of learning labs focused on facilitating dialogic, text-based discussions. Our research was spurred by the stubborn lack of classroom discussion in K-12 schools (Cazden, 2001; Nystrand et al., 2003)—a lack that has persisted despite the preponderance of empirical (Applebee et al., 2003; Engle & Conant, 2002; Kazemi & Stipek, 2001; Nystrand, 1997; Sedova et al., 2019) and theoretical (Bakhtin, 1981; Vygotsky, 1978) warrants for classroom discussions. We tracked conversational turns in videos of classroom discussions to track the frequency of student and teacher talk during our learning labs. This analysis revealed that teacher talk steadily decreased and student talk steadily increased over the course of a year of learning labs. There could be many possible explanations for these changes. One explanation is that the professional development supported teachers to change the way that they were facilitating discussions. However, the changes could have been a result of time of year—Perhaps students speak less during class discussions in the fall than they do in the spring. Another possible explanation is that during early learning labs students may have been intimidated to participate in discussions because there were more teachers in the room but, over time, they may have gotten more comfortable and spoken more frequently. The design of our study does not allow us to know the causes of changes in teachers’ discussion facilitation over time. That said, it was heartening to see these changes and we were interested to see whether there were any parallel changes in how teachers planned for discussions.
When we examined teachers’ planning, we focused particularly on the concept of responsiveness. We chose this focus in light of recent debates about PBTE, in which scholars had questioned, among other concerns, whether PBPDs could support teachers to be responsive to students and to context (Daniels & Varghese, 2020; Kennedy, 2016; Philip et al., 2019). This important critique of PBPD led us to ask whether, and in what ways, teachers in a PBPD planned to be responsive to students and context as they worked together on the core practice of facilitating discussion. Our investigation revealed that each consecutive learning lab saw an increase in the frequency with which teachers engaged in what we came to call responsive reasoning. Emerging out of our data analysis, we coined this term to describe a type of pedagogical reasoning that is grounded in an assumption that particular pedagogical decisions require situated contextual knowledge and therefore can only be made in practice. When responsive reasoning occurs during lesson planning, teachers engage with an anticipated pedagogical dilemma (e.g., how to follow a student’s idea as it emerges), use judgment to consider possible instructional options, and ultimately leave the dilemma unresolved until they can make a contextualized decision with students. While some might mistake responsive reasoning for being unprepared or “winging it,” it is quite the opposite. When teachers engage in responsive reasoning they make a concrete plan to do something that can only be specified in the specific context of its enactment. One example of this is making a concrete plan to engage in uptake of student ideas by using students’ ideas as resources in instruction (Collins, 1982). When engaging in responsive reasoning about uptake prior to instruction, teachers might set a goal to engage in uptake, they then might consider places in their lesson where they might engage in uptake, but they cannot finalize their uptake plan until they have students’ ideas on the table in the midst of a classroom discussion. They prepare to engage in uptake by reasoning through their lesson, but they leave that reasoning process with many possibilities open, rather than one specified plan. We contrast responsive reasoning with what we have come to call proactive reasoning, or pedagogical decision-making that teachers fully resolve in advance of knowing exactly what students will bring to a discussion (e.g., considering and deciding on texts, desk arrangement, etc.).
Comparing our two analyses of student talk and teacher pedagogical reasoning, we identified a mirror-image finding: that as teachers steadily increased their frequency of responsive reasoning there was a similar increase in student talk during classroom discussions. Thus, as teachers moved more toward responsive reasoning in their collaborative planning, they made more space for students’ ideas and engaged students in more talk during instruction. Perhaps more significant, however, are the implications of our findings for teacher education and professional development. While further research is certainly needed before we know whether and in what ways proactive and responsive reasoning are related to teachers’ discussion facilitation, simply delineating a difference between these two types of pedagogical reasoning has implications for teacher educators and professional development providers. In the following sections, we further interrogate the implications of our findings, both for practice and for theory.
Implications for Teacher Education Practice
Teacher educators, coaches, and mentors frequently engage teachers in lesson planning as they prepare for practice. In fact, Klette and colleagues’ (2017) recent study of opportunities to enact practice in teacher education programs in Finland, Norway, the United States, Chile, and Cuba found that, compared with other indicators of enactment (e.g., analyzing student work and rehearsing teaching), teachers had extensive opportunities to lesson plan. Interestingly, lesson planning is primarily proactive work. When planning, teachers make decisions before students enter the room to ensure that the time spent in class is productive and meaningful. Teacher educators, coaches, and mentors much less frequently spend time explicitly preparing teachers for aspects of instruction that are intentionally unplanned. Explicit preparation for the unplanned aspects of teaching may allow teachers to feel more prepared to do the work of making instructional decisions that are authentically in response to students. One body of research that has engaged extensively with this type of preparation for practice is research on approximations (Grossman et al., 2009; Kavanagh, Metz, et al., 2020). Research has less frequently engaged with how teacher educators might leverage opportunities to plan with teachers for the purpose of preparing them to be responsive in unplanned moments. Given how frequently teacher educators, coaches, and mentors plan with teachers, developing approaches to collaborative lesson planning that prioritize responsive reasoning may have significant implications for how teachers are supported to develop their practice.
Implications for Teacher Education Research
Our findings also have important implications for teacher education scholarship. Our analysis of teacher pedagogical reasoning led us to identify an important delineation between responsive and proactive reasoning. This distinction between two types of reasoning that teachers engage in while planning for instruction provides a more nuanced way for researchers to understand and conceptualize teacher sensemaking during instructional planning. Future research that explores this distinction would provide important understandings for how, and to what extent, teachers plan to be responsive to contextualized factors during instruction. In light of recent calls for teacher preparation to emphasize reasoning and judgment over prescribed or technocratic instruction (Daniels & Varghese, 2020; Zeichner, 2012), better understanding of how and to what extent teachers plan to be responsive in advance of instruction is critically important. Furthermore, while our findings indicate that the teachers in this study steadily engaged in more responsive reasoning over time, our current research does not speak to the causes of this observed shift. Research that explores this shift and the pedagogies of teacher educators that can support it would be a natural extension of this work.
Implications for Contemporary Scholarly Debates
Interestingly, while engaged in learning labs that were focused on a prespecified teaching practice, discussion facilitation, teachers became increasingly inclined to specify less of their activity prior to teaching. Many of the participating teachers had spent over 10 years in a large urban district with a history of employing top-down measures of ensuring that teachers follow stringent pacing guides and enact scripted curricula. For these teachers, becoming responsive to what students brought into the classroom on any given day was a counter-cultural proposition and in each consecutive lab, teachers became more comfortable being responsive to the uncertain circumstances of a dialogic classroom.
Contemporary and historical debates in research on teacher education have tended to set up binaries between pre-specified practices on one hand and responsiveness to students and to social context on the other. At times, the scholarship in our field, regardless of what side of these recurring debates it falls on, makes it seem as though the presence of one idea is harmful to the other: that pre-specified practices destroy responsiveness or vice versa. Our research pushes back on this imagined binary between prescriptiveness and responsiveness (Kavanagh, 2021). Instead, our findings point to the idea that the two might be in a symbiotic relationship with one another. In our case, at least, pre-specified practices sparked responsiveness, rather than detracted from it. When we make arguments for responsiveness and place those arguments in opposition to prescription, what we often fail to recognize is that without changes in our practice, all we can be responsive to is the status quo. All of us, teachers included, often need help in stepping outside of our traditions and sometimes that help can come in the form of clear first steps. Sometimes taking a few steps in a prescribed direction can change our position enough that it allows us to see what our students were capable of all along and until we see our students and hear their ideas, we cannot respond to them.
Footnotes
Acknowledgements
We thank Chenelle Boatswain, Lauren Szczesny, and the many teachers and leaders who have helped us imagine new possibilities for professional learning.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work has been supported by the James S. McDonnell Foundation. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the funders.
