Rethinking Collective Reflection in Teacher Professional Development

Abstract

Collective reflection, which has become a de rigueur activity in teacher training and professional development, is predicated upon Schön’s theory of reflective practice. This concept, according to which people learn to be reflective-in-action through reflection on practice, relates primarily to individual and one-on-one mentorship processes. The shift from individual to collective processes has gone largely unstudied and unproblematized. This study of collective teacher reflection in a professional development workshop calls prevailing assumptions into question by bringing an alternative lens, textual trajectories, to bear on this ubiquitous activity to better account for oft-ignored issues of context and identity. Using linguistic ethnographic methods, it traces textual trajectories of key ideas indexed in a collective reflection event. Key findings include the nonlinearity of the reflective process and the centrality of identity-work in collective teacher reflection. This study thus reveals functions of this ritual that belie its ostensible purposes and suggest a rethinking of this practice.

Collective reflection, which has become a de rigueur activity in teacher professional development (PD), is predicated upon Schön’s theory of reflective practice. While initial articulations of this theory (Schön, 1983, 1987) seem to suppose individual reflection, supported via one-on-one coaching and mentoring relationships, proponents of collective reflection seem to suppose that reflective interaction in the group setting fosters reflective practice by individuals (Jay & Johnson, 2002; Osterman & Kottkamp, 2004). This supposition has not been adequately examined or problematized. The current empirical study challenges these prevalent (though sometimes implicit) theoretical assumptions by analyzing teacher reflections at the end of a PD workshop both in their immediate context and along their textual trajectories—that is, the ways in which written and spoken text and other semiotic artifacts change and move over space and time (Lillis & Maybin, 2017).

Theoretical Frameworks

Reflective Practice

Schön advanced a theory of professional knowledge that accounts for professions characterized by tacit, embodied knowledge, rather than the types of knowledge typically studied—and privileged—in formal, academic classrooms. He critiqued the professional knowledge paradigm of technical rationality, a presupposed one-to-one correspondence between problems and their research-based solutions (Schön, 1992/2001, p. 3), for failing to take into account professions and situations that are more fluid and ambiguous than those accounted for by a positivistic scientific model. Education is categorized as such a profession, alongside, for instance, social work, divinity, and town planning. These professions, according to Schön, involve artistry in their practice; he seems to have used this term to counterbalance the “scientific” underpinnings of “major” professions such as law, medicine, and business, while the “minor” professions were perceived to offer a pale imitation, or science suffering from lack of rigor. Instead, Schön identified the need for “an epistemology of practice which takes full account of the competence practitioners sometimes display in situations of uncertainty, complexity, and uniqueness” (Schön, 1992/2001, p. 8). This kind of embodied knowing, which practitioners are not always able to theorize, or even explain or describe, entails recognition and judgment, which Schön termed reflection in action. According to this theory, when people’s routine actions produce surprising outcomes (i.e., something did not work as intended), they notice that something unexpected has happened, and through a reflective process, reformulate their understanding of the situation to decide what to do next.

Schön recognized some limitations of reflection in action, in particular that people may adhere to an unduly narrow set of possibilities for action. Analyzing the case of a town planner failing to convince a builder to undertake a project, he concluded that the planner focused upon his intermediary role and strategized about unilaterally controlling the situation, at the expense of greater transparency that would allow the builder to directly articulate his concerns and the planner, in turn, to address them. A priori commitments in realms such as the practitioner’s “role frame, strategies of action, relevant facts, and interpersonal theories of action” will limit their ability to reflect more broadly and, therefore, consider additional courses of action (Schön, 1983, p. 234).

One means of improving reflection in action is to reflect upon that process. Whereas reflection in action takes place during the event, when the outcome is still subject to the person’s actions, reflection on practice takes place after the fact. This reflection makes explicit the assumptions and commitment implicit in the action and subjects them to critical analysis (Schön, 1995, pp. 30–31).

This reflection on practice is presumed to impact future reflection in action. Thus, reflecting on practice is seen as a means of improving practice; the theory is that people can learn to be reflective in action by reflecting on practice with more experienced practitioners acting in a coaching capacity, in what Schön (1987) termed a reflective practicum. Indeed, within a decade of the 1983 publication of Schön’s The Reflective Practitioner: How Professionals Think in Action, reflective practice had become a much-celebrated gold standard of teaching, and consequently of teacher education (e.g., Adler, 1991; Bullough & Gitlin, 1989; Munby & Russell, 1989; Osterman, 1990; Ross, 1989; Roth, 1989; Zeichner, 1987).

The conceptualization and implementation of reflective practice in teaching and teacher education have come under fire for under-theorization and lack of theoretical and conceptual clarity (Beauchamp, 2006; Collin et al., 2013; Fendler, 2003; Hébert, 2015); hollowness when it becomes a forced ritual (Galea, 2012; Hobbs, 2007); and insufficient attention to issues of context and identity (Beauchamp, 2015, including an extensive review of prevalent critiques). Marcos et al. (2011), in a review of 122 articles relating to teacher reflection, found significant gaps between “what is said in research on reflection and what is communicated in practice” (p. 33).

These critiques are especially important in light of reflective practice shifting from an individual to a collective endeavor (Foong et al., 2018a, 2018b; Rantatalo & Karp, 2016), as this shift not only demands its own theorization but also heightens the abovementioned concerns with ritualized performances, context, and identity.

Indeed, the emphasis on reflection in preservice teacher preparation and especially inservice PD programs (which seldom feature close one-on-one mentoring relationships) all but guarantees that reflective processes will be undertaken collectively. The theory, implicit and even explicit in many practitioner guides (Hamilton et al., 2014; Jolly, 2004; Reed & Koliba, 2015), seems to be that people reflecting on practice together will ultimately be equipped to engage individually in reflection in action. However, this assumption—which requires several leaps in the realms of theory and practice—has not been sufficiently tested or problematized. This means that the widespread implementation of collective reflection in teacher PD, presumed to contribute to improvement of teacher practice, may not in fact accomplish this PD goal, and may entail—instead or in addition—undetected, and even undesirable, processes.

Collin and Karsenti (2011) pointed to, and attempted to rectify, the under-theorization of collective reflection. Their literature review of 52 documents yielded two predominant theoretical frameworks undergirding this body of research literature: Lave and Wenger’s (Lave & Wenger, 1991; Wenger, 1998) conceptualization of communities of practice (e.g., Allard et al., 2007; Buysse et al., 2003; Passman, 2002); and Vygotsky’s (1962) socio-cognitive approach to semiotic mediation (e.g., Korthagen, 2001; Pugach & Johnson, 1990; Reingold et al., 2008). Since the authors were primarily interested in the relationship between reflective practice and verbal interaction, they elected to focus on offering a more robust Vygotskian theorization. They essentially transposed Vygotsky’s model of semiotic mediation, whereby thinking is developed through verbal interaction in the social realm, onto teacher training. They recognized the leap required in the move from collective to individual practice without semiotic mediation, and attempted to account for it: “we hold that adults are sufficiently autonomous at the cognitive level to pursue the reflective process individually, in parallel to a reflective process that they develop through verbal interactions with each other” (p. 576). This supposition would seem to lie at the heart of the assumed connection between collective and individual reflection, but is not probed further or provided with warrants.

Another assertion in Collin and Karsenti’s articulation of a Vygotskian model calls to mind the above critique regarding the emptiness of ritualized reflective practice. They suppose that reflective language is an “indicator” (p. 578) of reflection, that is, that if people can talk about their reflection, or talk in the language of reflection, they are in fact engaging in reflection. Finally, Collin and Karsenti’s choice to de-emphasize social interaction (p. 573) evokes the above critique of research on reflection in general as failing to account for issues of context and identity.

The literature on teacher reflection reviewed above thus leaves undertheorized ways in which collective reflection can be understood using socio-cultural models of human interaction. However, if context and identity are continually and mutually co-constructed in interaction (verbal and nonverbal), the move from individual to collective reflection is far more complex than stated (and implicit) theories and prevalent practices would suggest.

This study identifies and exemplifies some shortcomings of prevailing assumptions about collective teacher reflection, through an empirical case study from an inservice teacher PD program. In particular, I analyze teacher reflective statements in the final meeting of a year-long PD workshop in a way that attempts to account for their functions and meanings both in their immediate context and against the broader workshop contexts. To trace these discourse events across contexts, I draw upon the concept of textual trajectories as an alternative lens for understanding collective teacher reflection as a discourse event.

Textual Trajectories

Recent work in sociolinguistics has expanded the focus “beyond the workings of language and text within specific events to the projection of language and text across them, in textual trajectories” (Blommaert & Rampton, 2016, p. 32). These trajectories are comprised of processes of entextualization and recontextualization, which are ongoing processes of meaning-making and transposition as texts (including varied communicative modes) are continuously reevaluated and altered across contexts (Blommaert & Rampton, 2016; Maybin, 2017). Contexts are interwoven rather than discrete entities, and simultaneously shape and are shaped by the utterances and texts enacted within them (Andrus, 2011).

The theory of textual trajectories, drawing upon Bakhtin’s notion that every utterance is in dialogue with past and anticipated future voices (Bakhtin, 1981, 1986), posits movement both forward and backward in time (Maybin, 2017), as each recontextualization responds to that which came before it, its immediate context, and future anticipated contexts. Maybin (2017) argued that the study of textual trajectories can contribute to “meso-level theorizing of textually mediated processes,” by helping to build “an explanation of how practices and agency at the micro level are dynamically interconnected with macro-level institutional processes and ideologies” (p. 416). Wortham and Rhodes (2015) similarly conceptualized textual trajectories as a bridge between micro and macro levels, advocating a new unit of analysis in the study of narratives: “chains of linked events, which together form a trajectory across which important functions of narrative are accomplished” (p. 160). Such analysis is predicated upon the understanding that social identities and positioning do not take place as discrete events in disconnected contexts but are formed and enacted in ongoing interaction across events and contexts (Bucholtz & Hall, 2005; Wortham, 2006). Thus, teacher reflection in a collective PD setting may be viewed as merely one event (i.e., a participant’s verbal reflection) in a linked chain, which extends through all the events and social contexts related to the broader collective setting (such as the same participant’s statements or actions in earlier workshop meetings). I, therefore, use a textual trajectories approach to investigate the myriad functions of this practice and problematize some of the basic assumptions that seem to render it so prevalent.

Research Setting

The broader context of the study is a nationwide initiative to develop in-school teacher pedagogical leadership. My university-based research team worked over a 7-year period in partnership with two large Israeli school districts to design, implement, and scale a program to train and support leading teachers, who convene weekly in-school meetings with their elementary school teacher teams. The leading teachers participate in bi-weekly PD workshops intended to foster the skills, tools, and dispositions to promote productive pedagogical discourse (Lefstein, Louie, et al., 2020; Lefstein, Vedder-Weiss, & Segal, 2020) in their teacher team meetings.

As the research-accompanied initiative progressed, it became apparent that while many teacher teams were leveraging the program tools and ideas to talk about general pedagogical issues; teaching and learning in specific disciplines were scarcely discussed (Segal, 2022). Similarly, district leadership often commented after site visits that the discussion was “superficial” and “general.” They, the research team, and the third-sector Research and Development unit responsible for the program at the national level concurred that discipline-specific teacher discourse—and concomitant teacher learning—required intervention.

This intervention took the form of a pilot program, in Year 6 of the leading teachers program, with 12 Language Arts leading teachers from six schools in one Israeli town. The town, located in Israel’s geographic and socio-economic periphery, was selected for this intervention by district leadership, whose stated considerations included serving a population as yet unreached by the leading teachers initiative and advancing an intervention geared toward all schools and teachers and not only elite ones. Potential participating schools were identified and recruited by the District Superintendents for Language Arts, and each principal selected the specific teachers (i.e., the 12 leading teachers participating in the PD). Stated (though not necessarily actual) acceptance criteria included identified leadership ability, a minimum of 5 years’ teaching experience, and openness to filming one’s classroom teaching. While the facilitation team conducted informational sessions as well as interviews with principals and their selected teachers, the potential pool was small and participants were essentially selected by the district. Participation was framed as a privilege and as an indication of the district’s and principals’ high regard for the teachers, who seemed glad to have been selected but did not necessarily know what to expect from the program. Most participating teachers, all of whom were female, had 8 to 15 years of teaching experience, with several outliers including a 20+-year veteran and a 3-year relative novice. All had taken part in previous inservice PD in Language Arts, and as part of the intervention, each participant led an in-school Language Arts teacher team, with each of the two participants ¹ per school charged with a range of grade levels across the 6-year elementary schools.

The bi-weekly PD workshops over the course of the 2019 to 2020 school year were jointly planned by representatives of the district (Yakira ² ), the national Research and Development (Gila), and the research team (Aliza) and facilitated primarily by Yakira and Gila. I mostly served as a researcher documenting the meetings but also occasionally facilitated portions of them. My researcher positionality entailed ongoing “cognitive oscillation” (Caronia, 2018) between insider and outsider perspectives. I had been involved in both the formulation of and research on the leading teachers initiative from its inception and was committed to its success. I was also critical of the lack of rigorous, discipline-specific teacher talk in the general teacher teams and brought to the Language Arts leading teacher PD my own ideas—research-based or otherwise—about both PD and Language Arts instruction. I was thus fully invested in the intensive planning of and reflection upon the PD workshop sessions, even as I was also committed to researching the facilitation team and the workshops. For me, tensions were more pronounced in the “discursive dances” (Aiello & Nero, 2019) in which I engaged with Gila and Yakira in navigating our differing approaches and stances than with the participants themselves, who seemed to know that I was a researcher (my ever-present notebook and recording devices could hardly be ignored!) but to also associate me with Yakira and Gila’s team (complaining on one occasion that they were expected to facilitate their teacher teams alone, whereas in the PD, “there are three of you!”).

The workshops, as well as participants’ in-school teacher team meetings, were structured around learning cycles intended to help participants probe disciplinary problems of practice by accessing disciplinary knowledge and analyzing representations of practice. Teachers would then arrive at potential courses of action, test them in their classrooms, and use resulting insights and questions to launch the next learning cycle. This pilot program highlighted challenges including the limited disciplinary knowledge base of elementary school language teachers, who are largely generalists; tensions between hierarchical canonical knowledge and locally constructed knowledge-in-practice; identity construction of participants as leading teachers; and many others, including the school shutdowns and move to remote learning occasioned by the COVID-19 pandemic beginning in March 2020. However, it was also apparent through workshop participation and our observations in select in-school teacher team meetings that participants were adopting and implementing—to varying degrees and in various ways (cf. Segal et al., 2018)—the routines, tools, and ideas espoused in the program.

Data Collection and Research Methods

The group convened for the year’s final meeting in mid-June, on the Zoom platform. In this meeting, each participant in turn reflected on her experiences with her team. These reflections, video-recorded through Zoom and transcribed, form the centerpiece of this article. Analysis of select excerpts from this meeting also references a wide variety of ethnographic sources collected throughout the year. These include video-recorded PD workshops; audio-recorded in-school teacher team meetings; WhatsApp and email communications between the facilitation/research team and the entire group, and between individual facilitators and participants; audio-recorded facilitation/research team planning and debrief sessions; and other observations and artifacts, which together comprise the data for this study. Note that the participants in the PD, and in this study, are the leading teachers, and the focus of the study is their reflective practice in the context of their PD, and not reflective practice with their in-school teacher teams.

Data reduction and analysis proceeded through several phases. First, I repeatedly viewed and read the transcript of the final meeting to get a sense of the topics addressed and the framings participants provided for their reflection. I noted program ideas that were repeatedly indexed, and selected one for close analysis. This concept, representations [of practice] (further discussed below), was selected for its centrality to the program and its relative prevalence in participant reflections. For the seven participant reflections indexing representations, I combed my detailed field notes of the PD sessions and in-school meetings for representation discourse events featuring the specific participant, and returned to those events in the video and audio recordings to trace their textual trajectories. I then analyzed these trajectories, using linguistic ethnographic micro-analytic methods, which entail close analysis of language combined with reflexive attention to positionality and context (Rampton, 2006; Snell & Lefstein, 2013). I assumed rationality on the part of the participants and tried to understand the implicit underpinnings of their actions (Rampton et al., 2016); this is true for all of my research but especially important to me to emphasize in this context, where it would be all too easy to fall prey to the genres of program or participant fidelity evaluation. Since space constraints do not permit the presentation of all seven cases, I have selected three that illustrate different types of trajectories. These include a participant who selectively adapted and (mis)appropriated program ideas, one whose robust understanding and buy-in faced local recontextualizations, and one who openly resisted program practices and demands; each used differing and complex discursive means to construct in the reflection session identities of professional competence through program compliance.

Early drafts of this article were read by several colleagues and discussed at two of our weekly research group meetings. From their familiarity with the language, the methodology, and the context, they found the linguistic ethnographic analysis compelling; as a result of this peer feedback, I removed one instance of what a colleague suggested was perhaps over-interpretation. I, in addition, discussed the findings with the other members of the facilitation team and with the two featured participants who made themselves available, and they found the analysis to be credible. One, Yonit, offered some more specific comments and insights, which have been integrated into the analysis and discussion.

Findings

Framing the Focal Event: Preparation for the Meeting

The entire final meeting of the PD workshop was devoted to reflection, for which participants were asked to prepare in advance. The framing of the event is significant to its interpretation, and, therefore, my analysis of this meeting begins with analysis of the advance instructions.

Gila sent the following text on the group WhatsApp, 6 days before the meeting:

In advance of the meeting, we invite each of you to present to us the work process that you led with your teams over the course of the year. Each teacher will have about 5-7 minutes to present the team in the manner of her choosing (PowerPoint presentation, bringing work product, pictures, oral narrative) relating to a number of points: (a) the learning cycles and the topics that the team dealt with (b) products, representations, and insights (c) reflection about the team’s work and about leading it, relating to successes, strengths, challenges, and questions. This is a celebratory opportunity to consider together the process that we have undergone this year ourselves and with our teams and to mark the learning processes. See you!

Gila’s instruction indicates objects of inquiry, to include “learning cycles,” “topics,” “products, representations and insights,” and “the team’s work.” These objects represent central features of the program’s ethos and methods. The first numbered list includes learning cycles, the method used to structure the inquiry through stages including defining an issue, selecting representations and evidence, analyzing representations and evidence, and drawing conclusions and planning anew. The term issue was introduced to denote problems of practice; these issues were to be formulated as questions, and adhere to specific criteria for productively framed problems of practice (Horn & Little, 2010; Zhang et al., 2011). Gila referred not to “issues” but to “topics,” a more general and less program-specific term used to denote the arena from which the problem of practice emerges. A program-specific term featured in Gila’s second numbered list is “representations,” short for representations of practice. Such representations, means of making classroom practice visible outside the class, have been identified as a key to generative teacher discourse, with an emphasis on transparent, specific representations (Horn & Kane, 2015; Little, 2002), and especially video (Pehmer et al., 2015; van Es & Sherin, 2002). Discussion of problems of practice based upon representations of practice was a lynchpin of both the national leading teachers initiative and our pilot Language Arts program. Throughout the year, we worked with participants using primarily video of classroom practice but also artifacts such as student work and participants’ lesson plans and materials (likely the “products” to which Gila refers). The term representations of practice was introduced in the context of the PD workshop and became, alongside issues, part of the institutional language (DeVault & McCoy, 2006) of the group, often shortened to representations.

Gila’s third numbered point also indexes key program ideas, particularly the reflective critical analysis suggested by “relating to” not only “successes” and “strengths” but also “challenges and questions.” The three lists in Gila’s written instruction suggest different activity types. The first entails mere reporting of broad topics the team discussed during the year. The third asks participants to analyze their teams’ work processes. The middle segment, well-aligned with program methods, seems to ask participants to bring the types of evidence that the group used throughout the year—“representations” and “products”—to support the analysis delineated in the third segment. However, the inclusion of “insights” on this list suggests telling about something, and not presenting objects for joint examination. Thus, while the contours of the instruction seem to suggest that participants bring representations of practice to the meeting, the communication could also be interpreted as a request to talk about representations of practice.

It is important to bear in mind that this PD workshop was devoted to participants’ work with their in-school teams, and, therefore, it is this work, and not their own classroom practice, that was the primary object of reflection. To examine this reflection, against the broader backdrop of the entire year’s activity, I have selected focal passages and events relating to one concept that was central to the PD and, ostensibly, to participants’ work with their teams, representations of practice.

As discussed earlier, the term “representations” was characteristic of and constructed as unique to the PD setting. In the meeting itself, although Gila had encouraged participants to bring evidence and to prepare a presentation in the format of their choosing, none brought any work product or pictures; two prepared PowerPoint presentations (only one was shown, due to technical difficulties) and one used facilitation cards we had distributed to punctuate her presentation, but the rest seemed to be speaking extemporaneously. However, although no participants presented representations of practice, nearly two-thirds of the speakers mentioned representations in their narratives. Analyzing the use of this concept and its textual trajectories can, therefore, serve to illustrate and elucidate the nature of participant reflective talk in a PD context.

Focal Passages: Discussing Representations of Practice

Of the 11 participants who spoke at the meeting, seven mentioned representations. The following analysis features the textual trajectories of Yonit, Dorinne, and Lotem.

Yonit:

“Analysis of Representations from the Reading Diagnostics”

Yonit related to representations while describing the topics addressed by her team over the course of the year:

So of course, like the majority, the topic of reading fluency was at the center. So if it was reading fluency, analysis of representations from the reading diagnostics. And then to try to understand what the source of the problem is, what the child didn’t understand. What the problem is with his reading. Does it rest on the emotional side, on the side that he, to understand, to analyze, and to try to understand where it comes from for the child, the lack of understanding. And then to try to understand to give him effective ways of improvement.

This presentation captured the collective process undertaken in the PD, and the explicit assumption—though not a requirement—was that participants would model upon this their first learning cycle with their teams. Reading fluency was the overarching topic for most of the workshop’s prepandemic portion. The first learning cycle involved representations of classroom practice in which students read aloud; disciplinary knowledge regarding types of mistakes that readers make, along with their linguistic classifications and possible causes; and ways of thinking about interventions aimed at addressing specific problems. Yonit’s opening, “So of course, like the majority,” reflects her awareness that her team’s activities were similar to the activities of others. Nonetheless, she described the process. Some aspects of this description could have been taken verbatim from the program facilitators, such as “analysis of representations”; “try to understand what the source of the problem is/to analyze, and to try to understand where it comes from for the child”; and “to give him effective ways of improvement.” However, other elements are less well-aligned with program aims, for instance, the “reading diagnostics” from which her “representations” are drawn.

The first assignment in the workshop was for participants to pay attention to their students’ oral reading through the lens of the disciplinary knowledge they had learned and in light of their experience in the workshop using it to analyze a representation of practice. They were asked to bring a “representation” of an unspecified type, to consciously observe and in some way document oral reading errors in their classrooms. One participant cited a diagnostic exercise she had done with a student, in which she timed the student reading and counted the errors. The district Language Arts Supervisor was present at that meeting, and indicated that this method was no longer recommended, describing the problems with such diagnostic tests and suggesting alternative ways of assessing oral reading proficiency. Later in the meeting, Yonit returned to the diagnostic tests and directly asked the Supervisor several questions about them. For instance, she pointed out that a student under pressure of the test, “for emotional reasons,” may not read a word accurately, but may be able to isolate and read it correctly afterward; would this count as being able to read the word? The Supervisor discussed the cognitive demands placed upon the child in each situation and other factors, but continued to express reservations regarding the diagnostic tests. Yonit pressed forward, asserting that “it doesn’t seem realistic for me to pay attention” to the particulars of student reading errors in a whole-class situation, “to notice and to write” them down, though perhaps this would be more implementable in “small groups and differentiated instruction” settings. Another participant, Karina, suggested that by listening to different students in whatever settings she deemed appropriate, Yonit could “build a portfolio” of student reading profiles. Gila interjected that the idea was to be attuned to student reading errors in the whole-class setting as well, “to note them in your head”—in contrast to Yonit’s assumption that she needed to write these down and Karina’s suggestion of collecting detailed notes on each student on an ad hoc basis. This comment represents the approach we took throughout this segment of the workshop, that paying attention to the reading errors most prevalent in their classrooms would enable teachers to devise interventions on a whole-class basis.

Yonit’s description of “analysis of representations from the reading diagnostics” thus flies in the face of program ideas and reflects her assumptions and preferences as expressed in the exchange that took place early in the workshop year. In addition to the fact that diagnostic tests were not considered a desirable representation of practice in this context (nor a desired practice, as presented by the Supervisor), they are also a representation that does not reveal much about the teacher’s practice, the type of representation emphasized throughout the PD workshop.

We twice asked participants to video themselves teaching—producing representations of teaching practice—and Yonit was amenable to having her video not only shown and analyzed at a workshop meeting, but also distributed to participants for use with their teacher teams and shown at a meeting with Ministry of Education policy-makers. This video featured a practice that Yonit devised to address reading fluency: She engaged her students in morphological unpacking, using cards with potentially difficult words from the text they were poised to read. The workshop meeting at which the video was shown forms another link in the chain of Yonit’s representations of practice textual trajectory.

Before the group’s viewing of the video, Yakira invited Yonit to give “a bit of information about the class.” Yonit described it as a “relatively complicated class, one of the most complicated in the school” and further framed the situation as “a first experience in front of the camera, I was very stressed, you can see that I’m not natural.” This kind of face-work, protection of one’s public image (Goffman, 1955), is extremely common in teacher discussion of classroom video; pre-emptive face-work allows the teacher to address and, therefore, mitigate the potential face threat entailed by revealing her practice (Vedder-Weiss et al., 2019). It is thus unsurprising that Yonit prefaced the video viewing by telling everyone how difficult the class was, as well as how “stressed” she felt, asserting that aspects of her teaching practice or demeanor, unspecified yet perceived by her as undesirable, were “not natural.” She thus signaled to her colleagues that she knew that what they were about to see was not her best performance, but the negative elements were also atypical for her.

Yonit further discussed her discomfort (“and it was really difficult for me to see it at first, I’ve already seen it three times or so, so I’ve already gotten used to myself”) and then assessed the video:

After the fact, when I saw it, I was enchanted by the students, by their answers, by the cooperation. It’s not that there isn’t a bit of chatter, it’s not a calm class . . . They were very very cute, and I was very teacherly (said at a lower tone, as if with chagrin).

Yonit constructed a complex line whereby the video wasn’t perfect (the students are chatty, she’s unnatural and teacherly) but the lesson went well (students are enchanting with their answers and their behavior). Once she offered this interpretation, it was difficult to frame the discussion as attending to a problem she presented, as she positioned herself as having no problem at all with the lesson other than her “teacherly” demeanor, and blamed this on the situation of being video-recorded. Participants in the ensuing discussion gravitated toward brainstorming other methods of addressing the goals of Yonit’s lesson, or talking about what she “should have done,” rather than analyzing the interaction within the lesson. Yakira pulled the discussion toward analysis of “opportunities” presented by Yonit’s students’ responses, ways in which they were and weren’t exploited by her, and considerations that could inform such decisions. In the facilitation team’s debrief of the meeting, there was an overwhelming sense that the meeting had not gone as anticipated. As Gila put it, “on the one hand, everyone tried to protect her (Yonit), and on the other hand, there were hurtful things that were said about what she should have done”; our conversation focused on facilitation moves that could help frame such discussions differently in the future.

Although discussion of Yonit’s video did not entirely reflect the program’s ethos of critical-yet-supportive analysis of representations of practice to probe problems of practice, it did reinforce her status within the group as a confident and capable classroom teacher. The representation of practice ended up being used as an example of purported best practice, rather than an aid to critical inquiry, with the video modeling an intervention that others could adopt. In the workshop context, participants inquired further about Yonit’s cards, and Yonit described and demonstrated the notes she had written on the backs of the cards to help guide her lesson. Tovit, the other participant from Yonit’s school, mentioned in her reflection the ways in which she and Yonit worked together, and in which the program impacted the entire school faculty: “We spoke in a unified language and dealt with difficulties that there were. And it started from the lower grades, and it reached all the classes, all the fluency and accuracy. Those cards that there were, like we all spoke the same language.” “Those cards,” the type of cards featured in Yonit’s video of her classroom practice, became a major takeaway from the process and the primary impact of her representation of practice.

When Yonit was called upon to reflect upon her team’s work, she included representations as a lynchpin of the program, but with a different meaning than that practiced throughout the year. Turning the representations into “representations of the reading diagnostics” enabled her to reference them and connect them to the topic under consideration (reading fluency) without the methodology surrounding representations within the program yet not practiced regarding her representation. In so doing, she maintained her status as one of the more advanced participants, who could articulate well what her team had done and the ways in which it adhered to the program. In recasting the representations and seemingly detailing the issues addressed by her team (“try to understand what the source of the problem is/to analyze, and to try to understand where it comes from for the child,” and “to give him effective ways of improvement”), Yonit presented what appears to be a coherent and rigorous program of teacher team work and on-the-job learning. However, the only learning cycles she described were the ones done directly in tandem with the workshop. When it came to specifying the source of a student’s reading fluency difficulties (“what the child didn’t understand”/“lack of understanding”), she enumerated only one option, “Does it rest on the emotional side,” which was featured in some of the program materials but never a focus of the analysis. But by recontextualizing the representations, she managed to performatively demonstrate program adherence and proficiency.

Dorinne:

“It was Knocking on the Door of Each of our Classrooms”

Dorinne, the only participant presenting a slide presentation, described her team’s work:

We met once a week, on Sundays, and like everyone we dealt with the topic of reading fluency through joint learning, and down the road representations from the field also began to join, which enriched and enhanced the learning. We dealt with a number of issues, when I look at the process, like when I was preparing the PowerPoint, and I wrote the issues that we dealt with, so then I truly saw the development of the team, within, also within the selection of issues. Because at the beginning it was, we really stuck to the things that I acquired at the workshop, and so that was some kind of very very important anchor. And little by little we truly succeeded in formulating the issues together, from the needs of the team. And it was really exciting and significant.

In contrast to Yonit, Dorinne related to representations as part of her team’s ongoing work, separate from topics and methods drawn directly from the PD workshop. She acknowledged that the initial issue addressed by her team was common to all the participants, and did not dwell on its contents: “like everyone we dealt with the topic of reading fluency through joint learning.” She highlighted the “development” of the team regarding independence in “formulating the issues together” rather than “sticking to” the “anchor” of workshop materials and issues. The bridge between workshop dependence and the team’s greater independence—which Dorinne characterized as “really exciting and significant”—is representations: “and down the road representations from the field also began to join, which enriched and enhanced the learning.” She later expanded upon this point:

At the stage when we began to really integrate representations from the classrooms there was a step up in the quality of the learning, in the quality of the work, in the quality of the discourse. Suddenly, it was something that felt as if it was knocking on the door of each of our classrooms. It no longer remained out there in theory. But really from within the classes, from within our needs, from within our activity.

Dorinne described a developmental progression whereby reaching the “stage when we began to really integrate representations from the classrooms”—that is, team members’ classrooms and not representations from classrooms outside the school—gave rise to a “step up” in “quality.” This found expression in “learning,” “work,” and “discourse,” all terms used within the program in varying ways regarding the teacher team meetings, with the underlying rationale that a certain type of discourse is generative for teacher learning with the objective of impacting teachers’ classroom practice, or work. Dorinne offered an unspecified “it” (which I interpret to mean the program, or its central activity) that was “knocking on the door of each of our classrooms” and “no longer remained out there in theory.” This depiction of the urgency of the team’s work and its bearing on practice rather than mere theory culminated with its relevance to the local school setting (“But really from within the classes, from within our needs, from within our activity”) and echoed the earlier characterization of the issues as emerging “from the needs of the team.”

Dorinne constructed a cogent narrative of her team’s development using key program concepts (“issues” and “representations”) as progress indicator and catalyst, respectively. Her emphasis on the responsiveness and relevance of program methods to local needs reflected the program’s ethos; it was a significant talking point in school recruitment, and one of our district partners frequently touted the “tailor-made suit” for each school. However, we often lamented that this aspect was not borne out in the Language Arts pilot program, which it seemed to us had become a traditional PD setting in which participants duplicated the PD activities in their own schools. It is thus noteworthy that Dorinne described this element as central to her team’s work.

Dorinne emerges through this analysis as a participant who embraced and implemented program ideas regarding representations of practice. Observations in her in-school team meetings yielded a somewhat more complex picture. On one occasion, the team planned an intervention, targeting intonation in oral reading, which several teachers would implement and video-record. This activity was well-aligned with program principles, and Dorinne’s team was one of the few to fully embrace such practices. However, the specific intervention was identical to one appearing in a representation brought by a different participant to the PD workshop, so this event illustrates the same migration of perceived best practices as Yonit’s cards (I cannot be certain that this is the source of the intervention, but it seems likely from the context). Of course, sharing and adapting teaching strategies is a positive thing, and I don’t intend to imply otherwise. However, if demonstrating pedagogical practices for wholesale adoption becomes the primary function of representations, this runs counter to the use of representations to probe emergent problems of practice. Dorinne, then, adopted the migrating-practices approach prevalent among workshop participants, but opened it to problematization by generating further representations featuring the migrated teaching strategy.

In Dorinne’s next in-school team meeting, one teacher, Noa, presented her video. The ensuing discussion was one of the more robust I have seen, hewing closely to the representation, attending to student and teacher perspectives, and featuring multiple voices and opinions. Dorinne focused and refocused the analysis and modeled types of things to notice and ask. However, there was also extensive face-work justifying the teacher’s practice, predominantly regarding classroom management. In addition, at no point was the intervention itself problematized. At the end of the discussion, Dorinne reframed the activity:

So this was basically a representation, I also asked Noa to bring it because it’s something that has really occupied the team . . . and it’s arisen at all grade levels, and this is an excellent opportunity to see it in the nitty gritty . . . Because it’s something that’s acute here in the team, so we’ll continue to deal with intonation and in the wake of Noa’s representation, I want to show you a practice that deals with intonation.

Dorinne here emphasized the topical relevance of intonation for the team, saying that it “has really occupied” them, has “arisen at all grade levels,” and is an “acute” problem. Noa’s representation was brought because of its topic, but in Dorinne’s telling, it did not actually shed light on that topic. It laid bare the issue “in the nitty gritty” but did not further the teachers’ understanding of how to address it, including the intervention that Noa implemented. Rather, Dorinne used the video as a springboard for what she wished to teach: “in the wake of Noa’s representation, I want to show you a practice that deals with intonation.”

Dorinne was consistent across contexts in her emphasis on the situatedness of teacher learning in local concerns and practices, and casted representations as embodiments of local practice attesting to the relevance of this learning. However, even as she and her team engaged in high-level analysis of representations, this recasting nullified some of their potential for learning.

Dorinne was not one of the most vocal workshop participants and was less occupied than others, such as Yonit, with demonstrating her competence and knowledge. When she spoke, she came across as knowledgeable, articulate, and attuned to and on-board with the program’s aims and methods. She was the only one whom Yakira, after a site visit in her coaching capacity, praised publicly for the work she was doing with her team, and she and the other participant from her school were generally viewed by the facilitation team as the pilot’s star participants. Analysis of her talk about and use of representations demonstrates that even the most competent participants who are most aligned with PD aims and practices may engage in subtle recontextualizations, which are only recognized through the tracing of their textual trajectories.

Lotem:

“They Brought Representations”

Lotem enthusiastically described teachers on her team sharing representations:

Basically my teachers on the team really really took it on. At the beginning, it was a bit difficult for them to accept the things. Little by little they understood that the devil isn’t so bad. They brought representations. Sometimes even before I asked, they came and they showed me things that they videoed, things that happened in their class. They sent us, we had a WhatsApp group, they shared all sorts of things there. Lots of times I was surprised that they were cooperating so much and that it like flowed.

Lotem described a teacher team who “brought representations,” sometimes unsolicited, “showed [her] things they videoed,” and “shared all sorts of things” on a “WhatsApp group.” Lotem celebrated the very fact of producing and sharing these representations; she said nothing about how they were used or how they related to or supported teachers’ learning. This emphasis on showing and sharing suggests that the representations were used more as demonstrations of classroom practice than as objects of critical inquiry. Framing the fact of sharing representations as “difficult” and “the devil,” Lotem expressed her “surprise” at the extent to which the teachers overcame the difficulties and “cooperat[ed]” with the endeavor.

This passage may be interpreted as a success story, a participant whose team exceeded her expectations and overcame difficulties to engage with representations. Some additional data points about Lotem’s program participation, along the textual trajectory of her interaction with representations, offer a different lens through which to interpret her reflection.

For the second workshop meeting, participants were asked to bring “representations,” or documentation of student reading errors, as discussed above in Yonit’s case. While other participants cited data or impressions of student reading performance, Lotem brought a video of a student reading aloud. Neither the student nor the teacher appeared on camera; Lotem had asked a student during recess to read aloud while she filmed, and she held her phone above the book so that the book could be seen in the video and the student’s voice heard. This representation was praised for its richness—that is, the fact that others could hear and characterize the student’s reading and not rely on a secondhand report. Furthermore, Lotem was praised for her initiative in producing such a representation, above and beyond the expectations of the facilitators, and the representation was prominently featured in the following workshop meeting.

Alongside Lotem’s positive experience with a representation that did not feature her teaching practice was her adamant objection to video-recording herself in her classroom. The first time that participants were asked to do so, she refused and threatened to leave the program over it. Indeed, not everyone provided a video, and Lotem managed to avoid it. After many participants did not submit their videos the second time around, the facilitators addressed this in the workshop. Yakira emphasized that the assignment was “not optional,” normalized the difficulties that teachers have with filming their lessons, and asserted that the way to deal with it was “not ignoring it” but turning to the facilitation team for assistance. Gila framed representations as “a tool for learning” whereby no one expected perfect, well-edited products: “it’s a way to observe reality in order to improve it.” Lotem stated that her classroom featured “a very, very difficult reality” with several students whom she termed “psychiatric” in their needs and behavior. She spoke at length, quickly and in an impassioned tone of voice, saying “I’m not comfortable with it” and indicating her unwillingness to be videoed—which persisted even as Gila tried to reassure her that this “difficult reality” was her reality, and that a video of her lesson could present opportunities to address her difficulties.

Lotem, in the final meeting, thus found herself in the difficult position of asserting her facility and success with representations even though throughout the year she had resisted producing and analyzing representations of her own practice. Her reflection sidestepped this problem in two ways. First, she discussed “the teachers on the team” and not her own practice. Saying that “it was a bit difficult for them” at first, but ultimately “they understood that the devil isn’t so bad,” Lotem constructed a learning trajectory for her team members that she could not credibly assert about herself. Second, she included as “representations” things that teachers would show her and send on their group WhatsApp, perhaps akin to the representation that she produced featuring a student’s voice and a book. She was thus able to celebrate the fact of her team’s representations of practice, without submitting to the principle of using representations to probe problems of practice within teacher team meetings.

Discussion and Conclusions

Hearkening back to Schön’s roots in Dewey (cf. Hébert, 2015), reflective thinking is defined as “active, persistent and careful consideration of any belief or supposed form of knowledge in light of the grounds that support it, and the further conclusions to which it tends” (Dewey, 1910, p. 6). In other words, reflective thinking involves drawing conclusions, making generalizations (Dewey, 1933), and using reflection on the past to propel thinking about the future—and ultimately, with Schön’s development of reflection in action, to shape future actions.

In each of the featured cases, the participant reflected upon her work with her team surrounding representations of practice. Lotem’s reflection featured the most explicit acknowledgment of difficulties, whereas Dorinne focused most specifically on the benefits and Yonit seemed to offer a neutral description of her team’s work. What these accounts share is a cogent construction of events whereby a central program method has been embraced and implemented by the participant and her team. This type of reflection upon the impact of PD on a teacher’s practice over a period of time is common in PD settings and bears potential, within Schön’s theorization, to foster teacher reflection in action. However, the textual trajectory analysis reveals other forces at play, as participants formed their narratives against the backdrops of Dorinne’s compliance yet subtle recontextualization, Yonit’s adaptations and (mis)appropriations of program tools and concepts, and Lotem’s downright rejection of them.

The individual meaning-making act of narrativizing one’s past practice can shape thinking about and action within future practice; perhaps, for instance, Lotem will embrace her team members’ realization that “the devil isn’t so bad” and document her own practice, Yonit will be able to consider learning cycles beyond replicating those from the PD, and Dorinne will begin the next school year with in-house representations from the start. The collective setting may foster further individual reflection, as Yonit might hear Lotem and consider the added benefits of representations provided by multiple team members, and Lotem, listening to Dorinne, might wonder whether her teachers’ WhatsApped representations indeed address their pressing pedagogical issues, and this individual reflection on practice may yield reflection in action. I do not have the data to directly address whether or not these imagined events, or any like them, took place, either during the reflective session or in participants’ subsequent work with their teams. However, the ways in which participants reflected on their past and current practice in the featured data did not entail the forward-thinking drawing of conclusions and generalizations necessary, according to proponents of Dewey and Schön, to propel reflection in action. The analysis I have presented here makes a case for an alternative set of interpretations challenging prevailing assumptions and suggesting that collective reflection may not achieve its purported benefits.

Let us return to two central critiques of reflective practice as it is conceptualized and implemented: (a) reflection is often reduced to a forced ritual, and as such, is hollow; (b) theorization of reflection does not sufficiently account for issues of context and identity. I suggested earlier, in the theoretical frameworks, that these issues are exacerbated when reflection is collective, undertaken in a group setting rather than one-on-one mentorship. Indeed, the meeting had a ritual quality. Participants were not given a choice whether to speak, but were nominated in the order in which they appeared on Yakira’s Zoom screen. The written framing of the event reflected the tension between its conceptualization as a time to reflectively consider “successes, strengths, challenges and questions” and as a “celebratory opportunity . . . to mark the learning processes.” The reflections have a performative quality, pledging allegiance to program ideals and attesting to their implementation. Yonit, during a session in which we focused together on her data segments and analysis, offered her own insights regarding the performative nature of the event. Acknowledging the routine nature of reflection in PD settings in which she has participated (“it happens in PD, you always do the reflection”), she asserted: “You want to say what you want them to hear.” She indicated awareness that she had selectively adopted and adapted the program’s concepts and practices, but thought that the reflection session was not an appropriate place to discuss this. Rather, “You think, what am I supposed to be saying here?,” and in Yonit’s view, she was not the only one engaging performatively: “It can be excruciating, five teachers already said the same thing.”

It is not that participants had only positive things to say, or that they did not speak their minds; in fact, many of them used the platform provided in this meeting to complain about the demands of the program. However, this became yet another part of the ritual, and given the ritualized, performative nature of the event, it is unsurprising that many of the contents, when explored in greater depth through their textual trajectories, rang hollow.

The importance of context and identity emerges even more strikingly. Participants were essentially asked to publicly account for and make sense of their professional practice, while positioning themselves as competent and program-compliant professionals. They engaged in sophisticated recasting of their textual trajectories to construct cogent, accurate, and socially acceptable narratives. I do not suggest that they perpetrated subterfuge, intentionally or unintentionally; analysis of the textual trajectories reveals complex sense-making processes that are coherent, rational, and show the participants to be active agents in bridging the gaps between the program’s espoused vision and their own dispositions and practices. The facilitation team bears no small amount of responsibility for the ritualized, performative, and ultimately hollow nature of the event. With very few facilitation moves beyond indicating whose turn it was in the serial-reflection round, it is difficult to know whether things could have gone differently had participants’ reflective accounts been probed and questioned. Similarly, our tepid and conciliatory approach to working with representations at key points in the workshop can explain some of the participants’ construction of this key concept; I do not have the data to examine to what extent teachers, in individual verbal interactions with the district coach, attempted to refine their understanding, and how these interactions impacted upon classroom practice.

As a postscript to this study, I can offer what I learned from and about Lotem, Yonit, and Dorinne in follow-up interviews conducted over a year after the conclusion of the pilot PD workshop. Lotem withdrew from the program, citing its overly demanding nature. This makes sense in light of the textual trajectory analysis, which reveals her lack of identification with—and resistance to—some of the program’s central practices, despite the positive spin featured in her reflection turn. Yonit, who continued to serve as a leading teacher at her school and participated in a second year of PD workshops with Yakira, described the ways in which her teacher team has flourished under her leadership, emphasizing the reproduction of teaching practices such as the cards she used in her video representation of practice—again, well-aligned with her textual trajectory throughout the first year and belied by her reflection. Dorinne relocated with her family and, therefore, did not continue with the program. She reported identification with its concepts and ideas but her new school was not part of the leading teachers program so her opportunities for implementation were limited; she also reported a strong sense of burnout (exacerbated by teaching under pandemic conditions) and that she was considering leaving the teaching profession. This postscript offers a perspective of time on the events described, but I wish to emphasize that this research does not purport to assess what the participants learned from the PD or how it ultimately impacted upon their practice. Rather, this research sheds light on the meanings and implications of the participating teachers’ reflections in light of the textual trajectories that preceded them. These limitations of this study certainly invite further research that explores teacher interactions in broader contexts and more tightly facilitated teacher reflection sessions.

The analysis in this study shows that identity-work within the particular social context is a key—and perhaps even the key—facet of the collective reflection activity. When teachers are occupied with constructing a performative narrative while engaging in complex social positioning, it is difficult to assume that they are simultaneously engaging in a linear process whereby collective reflection on practice promotes individual reflection in action. While the social context of this study was leading teachers tasked with leading in-school teacher teams, and not teachers focusing directly upon classroom practice, there is no reason to believe that the reflective process would be any more linear for classroom teachers in a PD setting. However, the particulars of the positioning and identity-work may differ, and this is also a worthy avenue for further study.

Collective teacher reflection has become a taken-for-granted aspect of many teacher PD settings. This is based on theoretical and practical assumptions about the potential—and desired—impact of this activity on teacher practice. This study offers an alternative framework for understanding these interactions, suggesting a reconsideration of the role that they are expected to play.