Membership Categorization Analysis for Education Policy

Abstract

Membership Categorization Analysis (MCA) is a language-based approach that provides unique insight into education policy research. At its most basic, MCA is concerned with identifying the patterns with which individuals or organizations develop categorizations. Such information can allow education policy researchers to hypothesize about the effectiveness of policy, adapt policies to the particular constraints of the culture, or assess how policies change cultural categorization. This chapter illustrates the usefulness of MCA for education policy. The chapter analyzes the way different educational organizations in Bangladesh with differing degrees of international connectedness construct the category teacher.

Introduction

This chapter presents the usefulness of Membership Categorization Analysis (MCA) as a methodological perspective for educational policy research. MCA is a language-based approach to deconstruct common categorizations (e.g. teachers, students, parents) present in any culture. This deconstruction allows researchers to better understand how a culture creates boundaries specifying the expected actions, characteristics, and relationships in a given category (e.g. “good teachers form partnerships with parents”). To demonstrate the usefulness of MCA, the chapter includes a case example examining the extent to which a preeminent global education reform discourse on teachers has been taken up by three different educational organizations in Bangladesh. The case example illustrates in depth how one might conduct an MCA study as well as the kinds of findings that are gained. Thus, the purpose of this chapter is to (a) present a review of MCA scholarship including its genesis and current uses in education policy research, and (b) illustrate the use of MCA in education policy research using the Bangladesh case study.

The chapter begins with a discussion of MCA and its origins with Harvey Sacks and Conversation Analysis (CA). This is followed by a review of the uses (current and prospective) of MCA in policy research. Thereafter, I draw on the MCA literature to indicate how researchers do MCA. The bulk of the chapter consists of the illustrative example using the Bangladesh-global education reform discourse example, followed by a summary and conclusion.

Overview of MCA

MCA was first developed in Harvey Sacks’ well-known lectures. Sacks, considered one of the original developers of CA, took interest in the way that everyday language constructs the world and patterns interactions. Analysis of suicide line phone calls led to Sacks’ research that contributed to the development CA methodology (Sacks 1992), a qualitative methodology that focuses on the structures and sequential nature of talk. CA differs from other language-based methodologies in a number of ways: the most important for the purposes of this chapter are the importance of sequential positioning for understanding language and CA’s essentially emic approach. CA traditionally uses as its data recorded, naturally occurring conversations, and its analysis focuses on word-by-word and line-by-line interactions (Hutchby and Wooffitt 2008). Each part of the interaction is working to accomplish some action; activities achieved in conversation include requests, repair, proposals, accusations, and complaints, to name a few (Hutchby and Wooffitt 2008). CA adopts a somewhat radical approach to language-based methodologies in that it confines itself solely to the words in the interaction. Thus, analysis of the language itself does not look through a particular critical lens or theoretical background; however, findings from CA analyses can be used to inform social life (ten Have 2007). MCA similarly draws on these key ideas.

In the same lectures in which Sacks developed CA, he also identified membership categorization as an important resource used in structuring language. At its most basic form, MCA is concerned with identifying the terms speakers use to refer to others and the language attached to those references (Schegloff 2007). For Sacks, categories and the process of categorization is a key feature in conversation and bring organization to the interaction (Housley 2002). A common example to teach this principle is shown in the following excerpt, “The baby cried. The mommy picked it up” (Schegloff 2007). Without needing to ask a question, someone reading this will intuit that the activity of crying is natural to the baby, the mommy is the mother of that baby, and that picking up the baby is the responsibility of mommy. Sacks argued that language is built on categories that exist in a commonsense network to those inside the language culture. Critically examining the network surrounding categories gives insight into how a population defines that category.

While some scholars have suggested that Sacks’ two offspring, CA and MCA, are distinctly separate or that MCA is the byproduct of initial thought and CA the more mature outcome, Housley and Fitzgerald (2002) described how Sacks emphasized both ideas in conjunction with each other. Thus, this chapter follows Stokoe’s (2012) call for a “systematic analysis of membership categories” and “track[ing] categorial concerns in the same way that CA has pursued sequential practices” (279).

MCA in Education Policy

My search for literature leveraging the use of MCA in education found few studies employing this methodology. After searching through education policy databases, CA and discourse-specific journals, and broad search platforms like Google Scholar, I found that only four studies combined MCA and educational policy. Two of these studies assessed the operationalization of Spain’s and France’s language policies in primary classrooms receiving immigrant students (Dooly and Unamuno 2009; Bonacina-Pugh 2012). Dooly and Unamuno used MCA to determine that while the policy document established using Catalan as a means of social cohesion and plurilingualism as a threat to that, teachers in the classroom oriented to plurilingualism as a resource to achieve their educational goals. Bonacina-Pugh similarly differentiated between the policy as a text mandating monolingualism in the class and the policy in action where students did categorization work to decide which language to use with their peers (i.e. those with shared English background spoke English, while a mixed language background resulted in French interaction). Neither case made explicit recommendations for addressing the disconnect discovered between the language policies and the policies in action.

The other two studies addressed very different areas. Thomas et al. (2015) identified a trend in Australian policy documents to emphasize the responsibility of parents in students’ academic achievement and analyzed parent-teacher interviews to see whether parents and teachers were constructing a similar parent category. They identified teachers and parents engaging in similar categorization and warned that accepting such categorization implicitly shifts accountability away from educational administrators and bureaucracies. (MacLure et al. 2012) considered the role early childhood education policy has on developing negative reputations among students. They argued that such policies as audits and baseline assessments, age-related statements and goals within policy frameworks, student case files, and collective instructional approaches ossifies categorization work relative to the “good child” rather than allowing further development. The authors recommended against early interventions, providing wider spectrums of developmental mapping and smaller group instruction to avoid prematurely establishing categorizations.

As these studies demonstrated, MCA provides a unique method of understanding the individual impact and lived experience of educational policies. Categorization work revealed how students and teachers circumvent language policy to achieve more pressing goals than those of the educational language policy. In the latter two cases (Thomas et al. 2015; MacLure et al. 2012), categorization work being done by policies was shown to be affecting how parents and students were being categorized on a local basis. In each case, the findings problematized the taken for granted, inviting policy makers to reflect on the goals of their policies and the various impacts they had on teachers and students.

As suggested above, MCA can identify the boundary within which a culture allows a category (e.g. teacher, student, parent) to act. These boundaries are the actions, characterizations, and relationships prescribed for a particular category. Schegloff (2007) described it in this way, “If an ostensible member of a category appears to contravene what is ‘known’ about members of that category, then people do not revise that knowledge, but see the person as ‘an exception’, ‘different’, or even a defective member” (469). This kind of information can be critical to policy makers and administrators. Attempts to develop a new policy may be in vain should a policy contravene the commonsense understanding of a key category. An example of this based on the data used below would be an international organization using financial aid leverage to pressure a local government to implement learner-centered policies without recognizing the boundaries of the teacher category defined by activities such as lecturing and testing. Thus, studying the categorization of teachers can provide useful insight to policy researchers.

Doing MCA

In regard to MCA, Sacks (1992) offered more general analytical understandings and constructs but gave relatively little specific guidance on how one might conduct a MCA research study. Synthesizing Sacks’ development of MCA and subsequent key analyses using MCA, Stokoe (2012) developed guiding principles and key concepts for doing MCA. These guiding principles provide novices to MCA or CA a means of pursuing such an analysis.

Stokoe (2012) identified ten key concepts in the membership categorization literature described below with examples drawn from the texts analyzed in the illustrative example. These are the constructs developed by Sacks and used by MCA scholars generally. All of these concepts may or may not always be used in a study, but they roughly form the universe from which to draw when doing an MCA study:

• 1. Membership categorization device: the collective category that binds categories together (e.g. the membership category device for principals, teachers, and teachers aids would be school employees).

• 2. Category-bound activities: activities that are linked in the text or interaction to categories (e.g. “The students (category) are supposed to follow (category-bound activity) their teachers”).

• 3. Category-tied predicates: descriptor or characteristic tied to a particular category (e.g. “Qualified (predicate) teachers (category) are essential (predicate)”) (Ministry of Education 2010, 57).

• 4. Standardized relational pairs: categories linked together by expected duties or obligations (e.g. student-teacher, “The students are supposed to follow their teachers”) (CAMPE 2011, 116).

• 5. Duplicative organization: categories bound together in a common purpose with set obligations like a team (e.g. parents and teachers in an “educator team”).

• 6. Positioned categories: categories that are linked by a hierarchical relationship where one category being described by the lower or higher hierarchical category provides unique meaning (e.g. local teacher, qualified teacher).

• 7. Category-activity puzzles: categories linked in an unusual sequence (e.g. “student-teacher”) to perform a particular action of setting the category apart or making jokes.

• 8. The economy rule: seeing a single category invoked is sufficient to understand the categorization work taking place (e.g. “the students follow their teachers”) (CAMPE 2011, 116).

• 9. The consistency rule: if two or more categories are next to each other and can be understood as belonging to a common membership categorization device, then they should be.

• 10. Viewer’s maxim: “if a member sees a category-bound activity being done, then, if one sees it being done by a member of a category to which the activity is bound, see it that way” (Sacks 1992, 221).

Stokoe (2012, 280) also provides practical guidance for how one might conduct an MCA study:

• 1. Collect discursive data. Data could be interactional, textual, or both (the case below exclusively uses textual data).

• 2. Build collections of explicit mentions of categories (e.g. teacher, student) and membership categorization devices (e.g. school employees).

• 3. Locate the sequential position of the category mentioned within the interaction or text.

• 4. Analyze the design and action orientation of the interaction or text surrounding the category mention.

• 5. Look for evidence of how parties orient to the category and build upon or resist the categorization.

Stokoe (2012) made a key point worth considering here: the key concepts invoked by the speaker or writer are made meaningful by the discursive act and not in a preconceived, decontextualized way. Thus, the MCA study should examine how the parties involved orient to categories and link them to particular activities, predicates, and relational pairs in the local context as opposed to the analyst’s.

Illustrative Case of MCA in Education Policy: Bangladesh and the Global Teacher Construct

Background of the Case Example

To illustrate the use of MCA in educational policy, I examine the extent to which a preeminent global discourse about teachers has penetrated different levels of the education system in Bangladesh. Over the past 10–20 years, one of the primary focuses of global education policy reform has been on the role of the teacher (Rotberg 2010; Paine et al. 2016). Paine et al. (2016) described a multifaceted discourse about teachers that has become particularly influential across the world:

• 1. Teachers are the fulcrum of the education system, the primary reason for inadequate results and the solution to any problem.

• 2. Teachers are defined in terms of deficits; they lack training, they lack accountability, they lack motivation, there aren’t enough of them.

• 3. Teachers should lead child-centered instruction and should not primarily function as dispensers of information to students.

• 4. Teachers ultimately produce student learning that can be measured in achievement tests.

Paine et al. (2016) described some useful concepts for understanding how this global discourse could spread. The first is the idea of “global flows” (719). Ideas flow through a variety of mechanisms across the world, penetrating places and competing with local norms and traditions. Second, “flows occur through spaces, but not all spaces are the same” (Paine et al., 719). The degree to which global ideas and policies influence locales differs depending on the political relationships and networks at hand.

Many global education policy researchers have identified international organizations and international aid as one of the key levers by which education policy has become globalized (Verger et al. 2012; Stromquist and Monkman 2014; Rizvi 2004). This is generally accomplished by either making funding conditional on reform, using persuasion and expertise to implement certain policies, or establishing common objectives that require policy harmonization (Verger et al. 2012).

The case of Bangladesh is particularly useful to examine the theories of global flows in spaces and the role of international organizations. Bangladesh is one of the largest recipients of international aid for education. Additionally, its educational system has both governmental and non-governmental schools with a significant proportion of primary education offered by non-governmental organizations (NGOs). Thus, my analysis is able to examine a variety of Paine et al.’s “spaces”. Specifically, this analysis considers the differential penetration of the preeminent global discourse in (a) Bangladesh’s Ministry of Education (MoE), (b) the Campaign for Popular Education (CAMPE) (a network of 1300 NGOs providing educational services), and (c) Dhaka Ahsania Mission (DAM) (an individual large, local NGO) using the insights gained from MCA.

Primary Education in Bangladesh

Around the time that the Bangladeshi Educational organizations indicated above produced these education sector strategy documents, primary education (classes 1–5) in Bangladesh continued to struggle. Bangladesh boasted one of the largest education systems in the world with 16.4 million primary-age children (all data from UNICEF 2009). While 90% of children were enrolled, only 51% of students completed primary education. Additionally, only 10% of schools reach the standard of 900 contact hours per year. Thus, many primary-age students drop out, and those who do remain receive relatively little instruction. Bangladesh does have a non-formal education sector that reaches out to youth who have dropped out to engage them in some amount of instruction.

The study posits that the selected organizations have different degrees of connection with the global education policy network which influences the degree to which they wittingly or not adopt the international teacher discourse described above. The MoE was the most connected because of its many direct interactions with international organizations and foreign aid agencies. In 2007, it partnered with over 20 different organizations on a variety of education system projects (Ministry of Primary and Mass Education 2007). CAMPE also connects with the international community but to a lesser extent than the MoE. It was founded after the 1990 Education for All declaration and defines its historical phases by its international partnerships (CAMPE 2014). CAMPE claimed key partnerships with three European foreign aid organizations during the timeframe in which the document included in this study were published, but previously had other international partners. DAM was founded in 1958 by a Bangladeshi Sufi. DAM engages in eight different primary education projects, two of which enjoy partnerships with an Asian foreign aid organization and an international organization, respectively. While these are only rough indicators of relative connection to international organizations, it seems reasonable to believe that organizations with more international partnerships would be more likely to engage with and potentially adopt global education trends than those that have fewer partnerships. Thus, I hypothesize that DAM was likely further removed from the global teacher discourse flow than CAMPE, CAMPE was likely more removed than the MoE, and that this will be apparent in the analyzed documents.

Case Example Data

For my analysis, I selected documents that would provide insight into how these organizations discursively develop the category of teacher. Also, given my own language constraints, the documents needed to be in English. Additionally, I wanted to find documents from approximately the same timeframe to capture the flow of the global discourse of interest at a particular time in the country. Documents from the same timeframe would ostensibly be responding to the same educational issues in the country. Therefore, for the MoE I selected National Education Policy 2010 (86 p.). This document outlines the policy priorities for the coming five years. For CAMPE, I chose Education Watch 2009–10: Exploring Low Performance in Education (121 p. excluding bibliography and appendices). CAMPE produces annual reports describing a particular issue or section in the Bangladesh education system. This particular report conducts an in-depth study of a particular region’s challenges in primary education to gain insight into the country as a whole. For DAM, I chose Education Sector Strategy 2009–2015 (37 p.). This is a straightforward document outlining its education sector strategy for the coming years. All of these documents build a teacher identity in the way they assign responsibilities and actions to teachers in their approach to the education sector.

These data contrast with the interactional or conversational data typically used in CA or MCA analysis. However, Stokoe’s (2012) key concepts and guiding principles of MCA opens up the data to be analyzed to include text. Additionally, the analytical features and purposes of MCA differ from CA in that they do not seek to identify sequential patterns and the intricacies of talk that require conversational data. Rather MCA focuses on deconstructing categorization work that is a feature of any kind of discourse while using sequential analyses. Thus, I argue that using MCA to analyze textual data is not incongruent with its intent.

Methodological Approach

For the purposes of this study, I focused on the following key constructs: membership categorization devices, membership categories, category-bound activities, category-tied predicates, standardized relational pairs, and the viewer’s maxim. The other means of analysis in MCA do not occur or only occur infrequently in the current dataset and will not be considered.

Using the key concepts discussed above to identify the key aspects of the dataset, I employed Stokoe’s five guiding principles. Thus, my dataset was built on the policy documents from the three organizations. I went through each text to build a collection of each instance of teacher. Each selection included the context surrounding the reference to capture the whole idea around the reference. Additionally, as I built collections, I erred on the side of over-including sequential exchanges to insure that all relevant initiating discussion was included. After building the collections and identifying the sequential positions of the categorial instances, I analyzed them based on the key concepts noted above. I read through the different selections of the data in each of the collections to identify common descriptions of category activities, predicates, and relational pairs, as well as highlighting divergent uses. I then analyzed each sequential segment to determine the policy document’s orientation to the category of interest. By examining the lexical choices surrounding the teacher category in each of these documents, this study demonstrates to what extent the dominant global “teacher” discourse has flowed into all levels of the education sector in Bangladesh. Thus, the case study both assesses the theory of global flows and how they impact local education organizations, and it illustrates how different organizations develop different frames for establishing what a teacher is and does.

Findings

To share my findings, I first share extracts from each of the education organizations and the line-by-line analysis that provides insights into how they’ve constructed “teacher”. These analyses are compared to the global discourse, as defined above, individually. Thereafter, I analyze the similarities and differences among the three organizations’ documents as well as document and compare the relative frequency of different categorical tools.

Ministry of Education

Toward the end of the policy, the MoE took up the topic of teacher dignity. This brief excerpt captured many of the key ideas that the education policy presents throughout the document.

• 1  If the issue of the status of teachers is limited to rhetoric and the teachers do not enjoy a

• 2  respectable social status in real terms, the quality of education cannot be improved. The

• 3  teachers are to be trained up as self-confident, efficient and important persons in the

• 4  cause of education. This is an urgent task. So, opportunities of training for them at home

• 5  and abroad will be created and stipends and training courses in the overseas countries will

• 6  be made available to them. These steps can strengthen the education sector. A separate

• 7  pay scale will be introduced for teachers of all levels to enhance their financial benefits.

• 8  (Ministry of Education 2010, 60)

Line 1 immediately attached the predicate “status” to teachers as a quality that teachers possess and later “enjoy”. However, ministry officials created a dichotomy of possible statuses teachers could possess or enjoy: a limited rhetorical status or a “respectable social status in real terms”. This suggests that the MoE automatically presents teachers with a social status, a rhetorical, respectable status simply conferred by being a teacher. This further suggests a teacher is an important, respectable societal institution. However, this status may not actually be enjoyed by teachers in their day-to-day practice. This issue was then tied to improving the quality of education in both lines 2 and 5, that is, teachers’ respectable social status is a necessary feature for education quality to improve. This first line indicated some features similar to Paine et al.’s construction of the teacher as the fulcrum of educational success. Teachers, through their status, were highlighted as a key element in improving quality education. However, the construction of teacher status linked to quality of education was not made as a necessary and sufficient condition which suggests that other factors may be critical in improving the quality of education.

The next sentence suggests that training is the remedy to teachers’ status being rhetorical or in “real terms”. Teachers currently lack training and are thus in need of it. The Ministry prescribed that the training is to make teachers “self-confident, efficient, and important persons in the cause of education”. This construction suggests that the reason teachers do not “enjoy a respectable social status in real terms” is because they have not claimed it through the manner by which they carry themselves. Self-confidence and self-importance are essentially ways of seeing oneself, which assumedly would come through in the teachers’ actions. Efficiency is ostensibly related to the way that the teachers conduct their work; however, this is not made explicit. Thus, the Ministry placed the responsibility on the teacher for the way in which the teacher is viewed in society and focuses primarily on self-belief rather than their skills or abilities. This construction builds upon the idea established in line 1 and strengthens the idea of teachers as the fulcrum of education. Considering a teacher’s social status only in regard to the teachers’ self-belief suggests a singular, even myopic focus on teachers that does not account for other stakeholders and conditions in the broader society. By not developing these broader considerations of social status, the MoE built a teacher construct similar in nature to the global teacher construct identified by Paine et al. and described at the beginning of the case study.

Additionally, lines 2 through 4 highlight the relative agency of teachers and the government in this training. The Ministry has constructed a problem where teachers lack social status due to their own incapacity in their person and their effort, and thus they prescribe training. Rather than allow teachers or local administrators the freedom to select the training most responsive to their needs, the government further prescribed what kind of training this must be. Line 4 provided an interesting construction saying that the opportunities for training “will be created” and stipends “will be made available”. However, these activities were not tied to any particular party. While the “will be” verb phrase suggests the most confidence in an activity being conducted, the lack of an organization attached to this phrase raises questions as to who will take on these important tasks. This focus on training and teachers’ deficits was the most common construction of teacher in the MoE document. This is also another aspect of the global teacher discourse, teachers defined by their lack of effective training.

Lines 5 and 6 are an interesting addendum to this discussion. After discussing teachers’ status and immediately linking it to a lack of training that would be addressed by some party in the future, the ministry officials indicated in the document that a pay scale will be “introduced for teachers…to enhance their financial benefits”. Since this sentence was included in a paragraph specifically focused on addressing teacher social status, using the viewer’s maxim rule, my analysis treats it as an answer to the lack of social status. What is interesting about this idea is that, unlike the above implementation of training, the MoE made no explicit connection of the introduction of a pay scale to achieving a change in teachers’ real status. It comes off as an afterthought, although it does address the relative social standing of the teacher in society.

In summary, the MoE developed a teacher construct that is deficient in training and status suggestive of the global teacher discourse. The Ministry began by describing a situation where teachers may enjoy a rhetorical or real respectable social status which is then confirmed as rhetorical. The responsibility for this was found in the teachers’ deficiencies and not a broader, more systematic concern. This is a teacher as fulcrum-type situation. Similarly, the answer to this was training which again aligns with the global teacher discourse. This teacher training was to occur but the Ministry did not specify by whom or when.

Campaign for Popular Education

The extract below came in CAMPE’s 2009–2010 annual report analyzing the factors inhibiting a particular region from achieving success in primary education. In the final section of the report, the authors summarized their findings and made recommendations for improvements. This excerpt captured much of the key categorical work done throughout the document.

• 1  The students are supposed to follow their teachers. If the teachers do not come to school

• 2  regularly or not be punctual, the students may do the same. If the students attend

• 3  regularly this is of little use because proper teaching-learning cannot happen without

• 4  simultaneous presence of both in the classroom. Such an environment has potential

• 5  negative effect on students’ attendance, teachers’ care of the students, quality learning in

• 6  the classrooms leading to ultimate dropping out of students from the system. We scanned

• 7  the meeting minutes of the school managing committees and, sadly did not find any

• 8  record of discussion on the issues. (CAMPE 2011, 116)

The first sentence captured a key idea in CAMPE’s construction of teacher, the critical relational pair between teacher and student. Students are bound to teachers by the expectation that they follow their teacher. This implicitly indicates that one of the teachers’ bound activities is to set an example. The following sentence is built on this by describing how this should be accomplished. The teacher, defined by their absence and lack of punctuality, was implicitly leading the student to do the same because of the teacher-student relational expectation of example-follower. This is something of a departure from the global teacher discourse described above. Instead of describing “teacher” by the student’s achievement outcomes, CAMPE discursively built “teacher” as successful in terms of the example set for the student.

The third sentence that runs from lines 2 to 3 returned to the categorial predicate frequently used by CAMPE of the teacher as not present. CAMPE often returned to the teachers’ presence because, as this sentence indicated, teacher and student presence was associated with “proper teaching-learning”. For CAMPE, when both parties were present, learning happened. However, when teachers were not present, teaching-learning could not happen. Teachers became a necessary cause for learning to take place, as did the student. This construction is interesting compared to the global discourse with which the study began. The teacher as fulcrum construction directs attention to the teacher, disregarding other parties. CAMPE’s construction of the teacher above indicates that teacher and student are equally necessary for learning to take place. This relational pair is activity-bound by learning as it could not take place independent of both parties.

The fourth sentence in lines 3, 4, and 5 turned the focus back on the teachers’ lack of presence and punctuality and ascribed negative consequences to the absent/late teacher. The subject of this sentence “such an environment” reached back to the previous sentence’s “proper teaching-learning” which was linked to the teachers’ hypothetical not showing up or showing up late. Thus, absent or late teachers were tied to decreasing student attendance, teachers’ care for students, quality learning in the classroom, and student dropout. This echoes the earlier idea of the teacher and student bound by an example-follower relationship: as teachers disengage from their role, students follow suit. To some extent, this negates the construction of the teacher described in the previous paragraph where both teachers and students were responsible for learning. In this construction, the teacher is the foundational element for student presence, learning, and persistence. However, CAMPE’s use of a distant chain to tie these predicates to the teacher seems to highlight their construction of a more networked teacher rather than a fulcrum-type teacher.

The final sentence spanning lines 5 through 7 indicated a new aspect of the CAMPE teacher as a teacher needing and lacking accountability. The School Managing Committee is a body of local community member and others responsible for the management and governance of the school (Al Mamun 2014). Part of their role is to ensure a functioning school environment. This sentence indicated that “the issues” of teacher presence and teacher punctuality were not addressed in any of the meetings. This is an obvious shortcoming, and its inclusion in this section establishes a link between the idea of teacher failure to fulfill responsibilities and the limited engagement of the school managing committee charged with governing the school’s functions. This tacitly suggests that were the school managing committee to discuss “these issues” and establish accountability for the teachers, the teachers’ actions would align with expectations. Thus, CAMPE situated its construct of teacher similarly to the global discourse as it sees a lack of accountability for teachers as a key constraint to a successful educational system.

In summary, CAMPE created a teacher that is tightly bound to the student by setting up an example-follower dynamic. Additionally, learning cannot happen without both teacher and student being present. Thus, the teacher is not a fulcrum in this system but part of a dyad where both parties are equally important. This CAMPE construction suggests an outcome orientation that is broader than the global discourse including responsibility, school attendance, and quality learning. Finally, CAMPE aligns with the global discourse of teachers lacking and needing accountability. Thus, CAMPE finds some common constructs with the global discourse while emphasizing other aspects that differentiate their construct of teacher.

Dhaka Ahsania Mission

Of all the education sector organization documents, DAM’s was the shortest and had the fewest references to the category “teacher”. Despite this, DAM developed a unique understanding of what it means to be “teacher”. DAM outlined different areas of focus for their education strategy. The section below was one of the specific objectives for their primary education goal of “creating an enabling environment for the quality learning of all children”.

• 1  Ensure community, local government institution members and teachers have

• 2  comprehensive understanding of issues of quality education in the target areas with

• 3  improved governance, accountability and capacity so that learning outcomes are

• 4  achieved. (DAM 2009, 26)

The first line develops a teacher relational pair that has not been encountered to this point. Teachers were bound to community and local government institution members. One of the categorical analytic tools developed by Sacks was the idea of a membership category device. A typical school category device groups teachers, principals, and students together. What DAM was constructing here is a school category device that link teachers, community members, and local government officials implicitly to students (whose learning outcomes were referred to) in the doing of education. Invoking the school category device, as opposed to other devices, seems clear due to the specific linking of these organizations to learning outcomes, one of the key purposes of schooling. This marks a clear difference from the global education discourse as described above. In that discourse, teachers were spotlighted as a party set apart from all others in its link to student learning. DAM however constructed a teacher bound to other parties, equally enlisted in the achievement of learning outcomes.

Some might argue that the collaboration DAM constructed here is part of a school governance device, rather than a school educational device, because of the discussion of governance and accountability in line 2. I dispute that because of the frequency with which DAM devises this idea of teachers, community members, local government institution members, and others being part of a collaboration that achieves learning outcomes. It seems clear from the document that DAM was intent on binding these to create a more holistic school device than is traditionally considered (i.e. teachers and students).

To ensure this collaboration achieved learning outcomes, DAM linked the collaboration to the need for “a comprehensive understanding of quality of education in the target areas with improved governance, accountability, and capacity”. Traditionally, one might expect individual parties in this collaboration to be bound by one of these activities (e.g. community and local government institutions’ members bound to teachers by governance, oversight, and accountability) or for teachers specifically to be singled out for developing capacity or needing accountability (as seen in the highlighted global teacher discourse and the MoE). However, DAM created a collaboration that collectively needed to have each of these category-bound predicates attached to them. Thus, governance, accountability, and capacity were shared responsibilities of the collaboration, each organization demanding it and providing it to the others in the collaboration. While this vocabulary of governance, accountability, and capacity suggests deficits similar to what Paine et al. describe in their review of the teacher discourse, DAM’s construction altered this idea. This deficit was not that of teachers but of the whole education ecosystem. All parties involved were complicit in this deficit situation, and thus all parties needed to partner to address the issue. This is markedly different from a system that describes external authorities imposing governance and accountability on teachers, and teachers attempting to avoid these through union protection or tenure rights (Paine et al., 755).

A note must be made about DAM’s tying of teachers, as part of a collaboration, to learning outcomes. As described above, the collaboration was repeatedly tied to this idea of learning outcomes. None of the instances specifically mentioned student-learning outcomes, but this may be assumed as the primary outcome of interest. This may initially appear to be synonymous with the global teacher discourse of tying teachers to student achievement measures. However, DAM never explicated how they operationalize the idea of learning outcomes. Additionally, leaning on the other Bangladesh education documents suggests a variety of possibilities of learning outcomes very different in nature than how the global discourse binds teachers and student achievement: student morality, student citizenship, student employment status. Thus, while the language may seem similar, the actual construct diverges significantly.

In summary, DAM discursively constructed a teacher as a part of a collaboration rather than as an individual. DAM described the collaboration as in need of governance, accountability, and capacity. This collaboration was responsible for learning outcomes, although it is not specified what these learning outcomes are. While these final two themes are reminiscent of the global teacher discourse, their development suggests significant differences. DAM attributed the lack of governance, accountability, and capacity to all of the parties in the collaboration, not only teachers. The possibility of learning outcomes meaning different things beyond the global discourse’s preoccupation with student achievement was also a significant differentiator. Thus, DAM built a very different teacher construct from the global teacher construct.

Comparisons Across Educational Organizations

Table 5.1 provides an initial sense of the different emphases, as measured by the number of times used, these educational organizations made on what it means to be a teacher. Listed under each organization are the top ten most frequently used predicates (i.e. descriptors attached to a category), activities (i.e. actions that categories do), or relational pairs (i.e. connections between the category and an external actor or object) connected to teacher. This table provides the opportunity to compare key ideas among the different organizations’ constructs, while also giving readers a more complete sense of the data I used to draw my conclusions. While this is not a traditional analytic step in MCA, I use it to provide greater transparency and to help readers understand the source of the patterns referenced below.

Table 5.1 Teacher categorial references by organization

The relative number of teacher categorical references is an interesting difference between the organizations’ documents and their categorization of teachers. The ratio is a more telling data point given the relative sizes of the documents. As noted at the beginning of this case, the global construct of teacher identified here centralizes the role of teacher, linking most educational activities to the teacher. The relative rate of references and variety of constructs attached to the teacher category illustrate the degree to which these Bangladeshi organizations approximate the global construct of teacher. Thus, the MoE invoked the teacher category 1.6 times per page, CAMPE did so 1.7 times per page, and DAM did so 0.6 per page. This comparison point suggests the relative importance of the teacher category in the organization’s education strategy as well as the complexity of the teacher category. So, DAM seems to formulate a less focal role for the teacher than the Ministry or CAMPE do. And, the complexity is vastly different as seen in the combined number of predicates, activities, and relational pairs found in each organization: MoE (53), CAMPE (64), DAM (18). These data suggest that CAMPE and MoE were closer in discourse to the global flow than DAM was.

The relative importance of relational pairs is another significant differentiator. While both CAMPE and DAM had relational pairs (teachers bound to students (16% of all the teacher category references in the CAMPE document) and teachers bound to community collaboration (35% of all the teacher category references in the DAM document) respectively) as their most frequent descriptor of teachers, that was less important for the MoE where their relational pair was the third most frequent descriptor (7% of all the teacher category references in the MoE document). This suggests that the teacher as constructed by the MoE is defined less by their interactions with others. This focus is reminiscent of the first indicator in Paine et al.’s description of the preeminent global discourse on teacher where the teacher is individually spotlighted as the primary party in education. The MoE deemphasized the role of other parties like students, parents, or community organizations in defining the teacher. In the same way, CAMPE and DAM departed from the global discourse by defining the teacher relative to others, diffusing the responsibility of education across many parties.

Another interesting difference relative to relational pairs is with whom the organization frequently tied their teacher. Both the MoE (7% of all teacher category references) and CAMPE (16% of all teacher category references) had teachers bound to students as the primary relational pair. DAM most frequent references included relational pairs with community collaborations (35% of all teacher category references) and students (9% of all teacher category references). As indicated above, DAM’s teacher construct suggests a much more networked and system-wide, holistic construct of teacher than either of the other organizations. This again denotes the difference between DAM and the global discourse that does not focus attention on the teacher as it relates to the community or other public organizations.

A key aspect of CAMPE’s construction of teacher as shown in the table above was that teachers are involved in the process of developing the teacher construct. Their report asked teachers questions and reported their answers, recommendations, and suggestions as seen in this example, “Teachers’ diligent presence in the classroom was seen by the teachers as prerequisite.” (96) Thus, teachers were established as agents sufficiently educated and aware as to provide insight. This was unique among the organization documents and the global discourse. In each of these other cases, teachers were not constructed as participants in developing the teacher discourse. Data are gathered from them, papers are written about them, but their voice and agency is not as keenly present as it was in CAMPE’s document.

An interesting observation is that with a few exceptions, ¹ none of the most frequently cited teacher category predicates or activities address what teachers do in the classroom. At the same time, all organizations defined teachers by their quantity or presence in school. This combination suggests that these organizations equate learning with the presence of a teacher. While each organization mentioned the idea of teachers as facilitators or child-centered learning once, the weight of their constructive efforts favored the presence of the teacher being sufficient for constructing an in-class doing of “teacher”. This rejects the global discourse of teacher above that defines teachers by their skill in making learning child-centered.

A counter-argument to this might suggest that the emphasis on training implicitly indicates a more robust in-class performance of “teacher”. This may be the case for the MoE, where one of the references to training indicated that training would be for “learner-centered pedagogical approach”. However, few of the other references tying teachers to training described what that training would look like. The few that did mention topics indicated training in self-confidence, information communication technology, or generic skills development.

Another significant absence from the organizations’ formulation of teacher was the role of student achievement in defining “teacher”. As indicated in the individual organization excerpts, some do mention the role of teachers in learning outcomes or quality education. However, in no instance were these ideas defined in terms of student academic achievement as measured through an assessment. This remains a teacher-bound predicate that has not been built into the Bangladesh education sector at this time, likely because Bangladesh had not participated in international achievement tests like Trends in International Mathematics and Science Study (TIMSS) or Programme for International Student Assessment (PISA) (and still has not). As Paine et al. described in their analysis, as countries engage in international student testing, their discourse around teachers begins to bind teachers to students in terms of a production function: teacher produce student learning as measured in achievement testing (2016). As indicated in Table 5.1, teachers were still defined in terms of inputs like teachers’ training and qualifications and in-class presence. Paine et al. indicate this is an approach from previous discourses (2016, 755).

Conclusion

This study conducted MCA on three Bangladeshi organizational documents to assess the degree to which different levels of educational organizations aligned with a preeminent global discourse defining “teacher”. These organizations differ in the extent to which they were connected to global organizations that might espouse this global discourse. Drawing on Paine et al.’s theories of global policy flows through spaces and networks, I hypothesized that organization documents from organizations with the most international organization and foreign aid agency partnerships would show the most evidence of a teacher construct similar to the global teacher discourse. The evidence presented above suggests that this is in fact the case.

In the example I provided in this chapter, MCA demonstrated distinct methodological insights to the question of global flows and local categorization. MCA indicated the extent to which categorization work done by the educational organization matched the preeminent global discourse around the teacher category. By using MCA, I was able to identify the different activities, attributes, and relationships that were expected of their teacher construct. While much of the categorization work used similar activities, attributes, and relationships, each organization emphasized different aspects, and these differences in emphases matched the global flows hypothesis. Of the three organizations, the MoE had the most international partnerships and aligned most with the global discourse on teacher. The Ministry showed a clear orientation toward an exclusive focus on the teacher as it deemphasized relational pairs with the teacher and, as shown in the excerpt, placed the onus on teachers for their social standing. Additionally, the most frequently used category description for the MoE focused on teachers’ deficits in training, another common theme in the global teacher discourse as indicated above. While the MoE did not tightly bind teachers to any particular pedagogy, they were the only organization to explicitly indicate the need for learner-centered instruction training.

On the other extreme, DAM had the least international partnerships and constructed a teacher category that differed in substantive ways from the global teacher discourse assessed here. In general, DAM deemphasized the construct of teacher as seen in its limited usage in their strategy document. Relatedly, it established a teacher as a member of a collaborative force rather than an individual actor. Teachers were bound to community members and local government institution members in achieving learning outcomes. DAM even transformed the concepts of governance, accountability, and capacity to make them characteristics that bind the collaboration together rather than something to be imposed on teachers defined by deficits.

CAMPE was in the middle in terms of both the number of international partnerships and its relation to the global teacher discourse. CAMPE distinguished itself by foregrounding the teacher-student relational pair and the teacher as an agent of influence in defining the construct. On the other hand, the focus on the teacher and the emphasis on teacher accountability suggested some similarities with the global teacher construct.

A final interesting finding is the ways in which the educational system discursively resisted the global teacher discourse described above. While some language was used to tie teachers to broad learning outcomes, these constructs did not approach the global discourse’s principal effort to tie teachers to students’ academic achievement. Additionally, the organizations did little to define the teacher in terms of performing “teacher” in the classroom. Whereas the global discourse creates a teacher who is child-centered and not an information dispenser, Bangladeshi education organizations built a teacher who was mostly present at which point learning could occur.

Summary

This chapter began by describing the MCA methodology. Developed concurrently with CA, MCA draws on CA’s attention to sequential positioning and emic perspective to analyze the component parts of commonsense categories occurring in discourse (e.g. teacher). This uncommon approach to educational policy analysis provides the possibility for researchers to understand how key categories of interest are bound by cultural category work and how this aligns with policy interests. The Bangladesh case illustrated how different levels of intercourse with global organizations was associated with differing degrees of global policy adoption as evidenced by common teacher attributes, activities, and relationships. Additionally, the absence of key attributes, activities, and relational pair characterizations demonstrated how local organizations resisted the global policy flow. Such findings could provide insights into the success of policy adoption and the ways policy can be adapted to effectively match local category boundaries.

Key Connections to Policy Research

• 1. Membership Categorization Analysis (MCA) is a novel method in educational policy research.

• 2. MCA can analyze key constructs in educational policy research using both interactional and textual data.

• 3. MCA deconstructs commonly used categories to identify the key activities, attributes, and relationships that define and bound the construct.

• 4. Understanding the construct’s categorization network allows policy researchers to hypothesize policy effectiveness, adapt policies to be effective in particular cultural categories, and assess impact of policies.

Note

• 1. Exceptions include the Ministry of Education’s “teaching capacity”, CAMPE’s “abuse students”, and DAM’s “facilitate class”. The first and the last are generic terms that provide no detail, while teachers’ physical and psychological abuse of students is a monstrous departure from actual teaching.