Individual Submission Summary
Share...

Direct link:

Stratified Representations: Textbook Choices and Symbolic Inequality in Brazilian Education

Sun, March 29, 4:30 to 5:45pm, Hilton, Floor: Fourth Floor - Tower 3, Union Square 1&2

Proposal

The promise of education lies in its ability to foster success for every student by creating an inclusive environment that values all backgrounds, experiences, and identities. As such, curricula and textbooks should embody principles of diversity to guide respectful and representative learning (Organizaci´on de las Naciones Unidas para la Educaci´on, la Ciencia y la Cultura, 2017; Cerna et al., 2021; Fuchs and Henne, 2018). However, textbooks also convey a “hidden curriculum” that shapes student perceptions, national identity, and societal norms (Jackson, 2002; Schissler, 2009). While Brazil’s national textbook program (PNLD)—one of the largest state-led procurement programs in the world—appears to create uniformity by providing all public schools access to an approved set of materials, the reality is more complex. In a country marked by sharp regional disparities, deep-seated racial inequalities, and persistent socio-economic divides, understanding how curricular materials are distributed is crucial. Schools and educational systems select works from within the approved PNLD list, and their choices may interact with these local factors to reinforce existing inequalities. To understand this relationship, this study analyzes the textbooks chosen by all public schools in the country and correlates those choices with the characteristics of the adopting schools. We first create a textbook profile by investigating how different social groups are represented, and then map the distribution of these materials to diagnose whether this policy is advancing equity or inadvertently reproducing patterns of exclusion.
This research treats textbooks not merely as content carriers, but as complex social artifacts where symbolic disputes are contested (Apple, 2013; Lerch et al., 2022). Our approach is guided by a multi faceted theoretical framework that explicitly connects content to deployment. We employ Critical Dis course Analysis (CDA) to deconstruct the content of the textbooks—the “what”—by uncovering hidden meanings and symbolic disputes embedded in language and imagery (Mullet, 2018; Moore, 2022). We then use the Policy Enactment framework to analyze the deployment of these materials—the “how”—by examining the school-level choices that bring a national policy to life and revealing the “invisible agen cies” that drive selection in practice (Ball et al., 2011). This dual lens allows us to connect the symbolic messages in the curriculum to their material distribution across the social structure. By linking the symbolic inequality present in textbooks to the material inequality reflected in school resources, we treat curriculum as a coupled problem, analyzing not only what books say but also where they go.
Recent advances in computational text analysis have enabled large-scale studies of textbook content, but this work remains largely constrained to English-language materials and often relies on selective samples, with a limited focus on low- and middle-income countries (Crawfurd et al., 2024; Voj´ıˇr and Rusek, 2019). Our research design overcomes these limitations. Unlike prior research, our study is built on a comprehensive dataset that includes not only the entire corpus of 102 PNLD-approved high school humanities textbooks but also the official school-level adoption records, granting us an unprecedented, nationwide view of curricular deployment. We developed a natural language processing (NLP) pipeline specifically adapted for Brazilian Portuguese to mitigate the “contextual loss” that occurs when applying generic tools to specific languages. This pipeline moves beyond simple word counts; for example, our dependency analysis examines the grammatical roles assigned to different groups, identifying whether women and Afro-Brazilian figures are positioned as active subjects of sentences or passive objects of actions. Similarly, our topic analysis reveals whether Indigenous groups are primarily discussed in his torical contexts or as contemporary actors in subjects like geography and sociology (Lucy et al., 2020). Furthermore, our analysis aims at incorporating multi-modal discourse analysis, examining how images and visual layouts work alongside text to reinforce or challenge stereotypes (Kress and Van Leeuwen, 2020). By developing a textbook profile considering its inclusiveness and combining it with municipal level statistics (income, HDI, etc.), we reconstruct a geography of adoption to test whether symbolic inequalities co-vary with material inequalities.
Our preliminary results show that across the corpus, narratives remain dominated by white, male, and urban figures, with women, Afro-Brazilian, and Indigenous actors often confined to supporting or
historicized roles rather than being presented as contemporary protagonists. While significant variation exists among individual titles, a clear, stratified pattern of adoption emerges. Schools in urban, higher income, and better-resourced municipalities consistently select more inclusive textbooks. Conversely, schools in rural, poorer, and racially marginalized municipalities disproportionately adopt books with narrower representational breadth. This means students in contexts already marked by structural disad vantage also face a symbolic curriculum that is less reflective of their identities, reinforcing a cycle where material and symbolic inequality converge.
This study offers a new methodological framework for studying educational equity by treating curriculum as a problem of both content and deployment. For policy, the findings provide actionable insights for the PNLD, suggesting that simply approving a diverse range of books is insufficient. Policymakers must also consider how adoption incentives and teacher training can ensure the equitable distribution of these critical symbolic resources. Ultimately, this research underscores that genuine educational inclu sion requires not only physical access to resources but also symbolic access to narratives of recognition, belonging, and identity.

Authors