The content of school textbooks is of perennial interest to educational scholars, teachers, parents and the broader public alike. This may be partly because textbooks are the tip of the iceberg, the most visible evidence of an education system that is vast and impenetrable. In many countries, students bring them home from school and, consequently, they serve as partial evidence of what goes on there. They are often viewed as the source of ‘official knowledge’ that a nation seeks to convey to its youngest citizens (Ngo, 2014). For these reasons, there is a multitude of textbook studies that examine what they have to say to their captive audience of student readers and how they say it.

Maria Grever and Tina van der Vlies (2017: 288) provide a useful definition of history textbooks, which they describe as:

Tony Taylor and Stuart Macintyre (2017: 620) have suggested that ‘the primacy of the textbook as a resource may soon be at an end.’ This point is contentious. We do know that in many nations the history textbook remains central to instruction (Kim et al., 2013; Naseem, 2014; Nishino, 2010). However, regardless of whether it is central or peripheral to instruction, it is an artefact of a society and how it seeks to remember itself (Morgan and Henning, 2013). It is critical that scholars are attentive to studying textbooks. It follows, then, that scholars must also be attentive to how they go about this work.

Eckhardt Fuchs (2011: 17) has asserted that the field of textbook research is mired with studies that take the research object, the textbook, as ‘its point of departure’ rather than ‘various thematic, methodological, and disciplinary contexts’. This assertion was a point of departure for our research team on the project Thinking Historically for Canada’s Future (THFCF), a pan-Canadian research endeavour funded by the Social Sciences and Humanities Research Council of Canada. Our Curriculum and Resources Cluster within this project has a mandate to examine the curriculum and resources, including textbooks, for each of Canada’s 13 provincial and territorial educational jurisdictions. As a means towards preparing for our own analysis of textbooks, we conducted a comprehensive analysis of other history textbook studies. The primary objective of our analysis was to understand what methods scholars selected for their textbook studies, why they selected those methods, and the strengths and limitations of their methods. Among our various findings is that scholars of textbook studies do not typically address their research methods in explicit ways (for example, Foster and Karayianni, 2018; Forget and Payanotova, 2003; Pingel, 2010). The lack of attention to explicit methods is troubling for the replication and relevancy of research findings. But, perhaps more importantly, this lack of attention to methods may well underpin our finding that scholars rely on a very narrow scope of research methods. We found that scholars rely overwhelmingly on content or discourse analysis, without attention to visuals, didactics, sociopolitical context and other complex aspects. Our findings lead us to agree with Ehaab D. Abdou’s (2018: 482) claim that there is ‘a clear absence of a holistic and integrated framework’ in textbook analysis.

To our knowledge, Grever and Van der Vlies (2017) were the first scholars to note the need for a holistic approach to textbook study. They were joined by scholars such as Abdou (2018) and only a handful of others to take note of holistic analysis. Grever and Van der Vlies (2017: 294) define holistic analysis as an approach that ‘goes beyond a representational analysis limited to one person, event, or period, separate from the context of the whole textbook’. For them, a holistic approach is an examination of the whole history textbook and not selected chapters, and is inclusive of ‘composition, periodization, [and] visual intertextuality’ (Grever and Van der Vlies, 2017: 286). While we agree that ideally a holistic analysis addresses the textbook in its entirety, we note that large comparative studies of texts often require selective samplings for analysis. However, sampling does not negate analysis beyond simply content or discourse. Instead, we call for a holistic analysis as both a qualitative and quantitative assessment of the complex features of a text, either in its parts or as a whole, which includes content, discourses, visuals, didactics and sociopolitical context, among other aspects. Using this definition, this article draws on our comprehensive analysis of textbook studies to demonstrate why it is beneficial for scholars to expand beyond a singular method and, wherever possible, draw upon a range of methodological approaches. In so doing, we seek to extend the yet to be taken-up call by Grever and Van der Vlies (2017) for a holistic approach to textbook studies.

We gathered a corpus of 104 peer-reviewed scholarly journal articles and book chapters published between 1991 and 2021 in English and French. Our sampling of publications was created by conducting searches in our institutional library repositories for history textbook studies over the past three decades. Our extensive list of studies is drawn from some of the most prominent social studies, history education, curriculum studies and media studies scholarly journals and prominent edited collections that offer major contributions to the field of history textbook analysis. To demonstrate the range of articles consulted, our corpus features work from the following 27 journals:

Table 1 shows our selection of 12 edited volumes. We included studies on a wide range of topics, including representations of Indigenous peoples, class, race, gender and human rights, and concepts of nation, nationhood and nationalism. We have representation from a range of nations across the world, including Canada, the United States, England, Australia and various nations in continental Europe and Asia. Our data set was not intended to be exhaustive, but rather to offer wide-ranging examples of relevant literature. Our aim was to scan the field broadly to identify current trends in studies of history textbooks, and more particularly to assess the methods employed for those studies.

We note that there are limitations to our data and the claims we can make on an international scale. For example, our focus is on studies of elementary and high school textbooks (Kindergarten to Grade 12 in most of North America, or Secondary IV in the Canadian province of Quebec and A levels in Britain), rather than those used in postsecondary classes. This is because the content of elementary and high school texts is typically vetted by educational authorities for wide use in classrooms, often across a nation. Postsecondary textbooks are far more idiosyncratic, since they are selected, and often authored by, individual instructors or instructor collectives. In addition, effort was made to generate a list of studies that was not isolated to the Global North. We acknowledge, however, that we do include a disproportionate number of studies from the Global North, a factor of language barriers and the dominance of English-language academic journals.

The process of gathering and analysing the data occurred from autumn 2019 to June 2022. We had developed a framework for the analysis of each textbook study that was approved by the THFCF Executive Committee before the study began. The framework required that researchers read articles deductively (that is, from general theory to specific findings) to critically assess epistemological claims, theoretical framework, research methods, application of methods, core methodological assumptions, strengths and limitations of the methods employed and the findings. In addition, researchers judged each article on its overall strengths and limitations, as well as on its contribution to the field. Due to the heterogeneity of each textbook study’s description of methods, or lack thereof, we could not conduct a meta-analysis that provides definitive quantitative pooling of results. Rather, given that researchers needed to assess the implicit methods used in some textbook studies, we provide a qualitative narrative description of findings. As such, after analysing all 104 individual studies, the research team identified prevalent trends across the data set. Researchers created ‘deep dive’ briefs to draw comparisons across the data set regarding these trends. Throughout the process, researchers shared and compared their assessments to maintain consistency, validity and reliability.

Our core research team, the authors of this article, engaged in discussion until consensus was reached regarding four areas that require greater attention in history textbook studies. First, textbook researchers need to be more explicit about their methodologies and methods. Second, there is a paucity of studies related to classroom use. Third, studies focus on content in isolation from the sociopolitical landscape of production (see also, by the same team, Clark et al., 2024). Fourth, studies tend to focus on one method – often an implicit use of content or discourse analysis – without paying attention to other aspects of the textbooks under study. For example, we found that visual and didactic analysis were rarely considered in textbook studies. We contend in this article that an explicit, holistic approach that is attentive to various aspects of the textbooks under study, inclusive of content, discourse, visual, didactic and more, is critical to advancing the state of history education resources. We are conscious of how ambitious and time consuming such a five-pronged holistic analysis is for a research team, let alone an individual scholar. We argue nonetheless that striving for a holistic analysis of textbooks encourages researchers to expand their range of methods and offers a far richer understanding of what and how students learn about the past. This article supports those few scholars already calling for holistic analysis of textbooks by offering an examination of where the field stands in its methods. To do so, we focus on the implications of narrow/singular versus holistic methods for the future of textbook analyses.

We found from our study that content analysis remains dominant. It is positivist in nature, and it often involves empirical analysis of textbooks with respect to war narratives (for example, Guichard, 2018; Klymenko, 2016; McCormack, 2000; Nozaki, 2008; Sharp, 2014; Roldán Vera and Fuchs, 2018) and narratives of national identity (for example, Carrier, 2018; Heathorn, 2000; Jackson, 2018; Schissler and Soysal, 2005). We also found that increasingly scholars are turning to discourse in addition to content analysis to focus on representations or omissions of sociocultural categories such as gender (for example, Woyshner and Schocker, 2015), race (for example, Crawford, 2013; King and Simmons, 2018), disability (for example, Sleeter and Grant, 1991) and LGBTQI+ people (for example, Höhne and Heerdegen, 2018; Schmidt, 2014). Content and discourse analysis often focus on selective text, to the exclusion of other important aspects of the textbook, such as visual and didactic analysis, and to the fuller context of the textbook’s production and national context (Choppin, 1992; Kress and Van Leeuwen, 2006; Nicholls, 2006b). Here, we address four of the most prominent methods: content, discourse, visual and didactic. Due to space limitations, we do not address textbook production in this article. This is an important element of a holistic approach, and it is an aspect we address in additional publications.

A holistic analysis is an attempt to expand the field from a sole focus on the written narrative of the textbook to other aspects (Repoussi and Tutiaux-Guillon, 2010). A holistic analysis is attentive to the complex aspects of the textbook and to its perception (that is, how it may be interpreted or used), rather than isolating for only one component. But scholars continue to focus on what is written, which explains the dominance of content analysis in the field of textbook studies. There is no doubt that the amount of content that publishers afford to a specific historical subject is critical to understanding what narratives are sanctioned by a state and, by virtue of this, whether we are making progress in history education. For example, the Ontario Black Historical Society in Canada created a 2020 public campaign #BlackedOutHistory to demand more Black history in schools. The Society showed that only 13 pages of a 255-page Grade 8 Canadian history textbook by Nelson Education contained Black historical content (Henry, 2020).

Content analysis alone, however, does not offer an understanding of how Black history is represented when included. In this case, a critical discourse analysis is necessary to interrogate the politics of representation – language choices and their potential ideological effects. Scholars are increasingly analysing texts for the ways social formations of power shape discourses, such as descriptors, word couplings or absences. For example, it matters if Canadian history textbooks address the institution of slavery by using the language of enslavement to recognise that Blacks were enslaved by Whites due to White supremacy, and not defined as slaves, or, worse, as indentured servants, as an identity marker that erases White supremacy.

But an analysis of what and how a text is written is incomplete without attention to the many other ways that history textbooks teach students about the past. Visuals, for example, are not frills meant to draw students to the real historical content of the text. Instead, they are fundamental sources of information that, when used well, can engage students in critical inquiries about historical evidence and interpretations. The importance of visuals for history textbooks is perhaps clearest when controversy emerges. In Civil Rights, Culture Wars, Charles W. Eagles (2017) describes the saga of the high school history textbook Mississippi: Conflict and change authored by James Loewen and Charles Sallis. This textbook, which the authors describe as ‘boldly revisionist’ (Loewen and Sallis, 1980: 87, as cited by Eagles, 2017), was rejected by the Mississippi State Textbook Purchasing Board in part due to the inclusion of a photograph of a lynching and the desire to have students ‘take pride in state history, not question it’ (Loewen and Sallis, 1980: 184, as cited by Eagles, 2017). This example is not intended to weigh into the appropriateness of this photograph, but rather to illustrate that visuals – inclusions, types, aesthetics – are integral to the constructed historical narratives conveyed to students in textbooks.

Like visuals, inattentiveness to didactics serves as a detriment to the scholarly pursuit of understanding what and how students learn from history textbooks. The United States’ history textbook controversies are again rife in this regard. For example, in 2018, Grade 8 students in Texas were asked to list the positive and negative aspects of slavery based on an activity from the 2005 textbook A History of the United States (published by Prentice Hall), which also argued that not all slaveowners were cruel (Greenlee, 2019). This example demonstrates the importance of asking whether textbook didactics ask for basic comprehension of historical facts (for example, repetition, two-sided debates and simplistic questions) or the cultivation of historical thinking (for example, evidentiary inquiries, significance, and cause and consequence). The latter is sorely lacking, according to the report Teaching Hard History: American slavery, conducted by the Southern Poverty Law Commission (2018), which included the analysis of 12 history textbooks. One of their findings was that most textbooks did not make use of original documents to ‘gain insight into some of what enslaving and enslaved Americans created, thought, aspired to, and desired’ (Southern Poverty Law Commission, 2018: 39). The report calls for textbook authors to ‘expand their repertoire of historical documents beyond the usual narratives to do a better job of representing the diverse voices and experiences of enslaved persons’, and, in so doing, ‘make intentional connections – good and bad – to the present, by showing the lasting contributions of African cultures and ideas, as well as the enduring impact of racial oppression on contemporary American life’ (Southern Poverty Law Commission, 2018: 41).

Of course, even with the addition of more methods of textbook analysis – moving from attention to only what is written to the inclusion of discourse, visuals and didactics – studies do not fully account for all possible aspects of a holistic approach. In addition to what we have shown in this article, the field of history textbook study still lacks consideration of the sociopolitical context of textbook production (that is, national context, authorship, publishers, approval processes) and the pedagogical use and student reception of textbooks in the classroom. For example, we know within the United States and Canada that Black authorship of history textbooks is rare, and that the teachers who use textbooks are overwhelmingly White (Greenlee, 2019). A holistic approach to textbook analysis may seek to examine the relationship between race, authorship and pedagogy.

We recognise that a holistic approach is an expansive research agenda if we are to include content, discourse, visual, didactic, sociopolitical context, usage and more in the analysis of textbooks. In our sample of 104 textbook studies, we identified five projects that managed to deploy what we have described as a holistic approach, with varying degrees of attention to the various methods and varying sampling sizes of textbooks (Araújo and Rodríguez Maeso, 2012; Carleton, 2011; Georg Eckert Institute, 2015; Mirkovic and Crawford, 2003; Sleeter and Grant, 1991). A study of Portuguese textbooks by Araújo and Rodríguez Maeso (2012) managed to check all the boxes of a holistic framework. By looking at the content and its discourse, they demonstrated that textbooks naturalise colonialism, slavery, racism and the idea of European union. Their visual analysis showed how Blackness was portrayed as backwards and needing of Western humanitarian intervention. Their attention to the didactic portions of the book highlighted how textbooks encourage empathy with enslaved peoples and their racialised descendants, yet fail to elicit critiques of the overall colonial-capitalist system that engendered these unjust relations of power. Finally, the authors, albeit somewhat briefly, discuss how political considerations and current events influenced the preparation of the book during the fifth centenary of the ‘Discoveries’. The goal was to project the Portuguese nation as one of ‘cultural mediators’, and their colonial enterprise as an ‘intercultural dialogue’ that helped bring about globalisation.

Unlike Araújo and Rodríguez Maeso (2012), Mirkovic and Crawford’s (2003) cross-cultural analysis of teaching history in Serbian and British schools is greatly imbued with sociopolitical discussions of the complicity of textbooks in producing and reinforcing politicised narratives of national identity. The authors begin ‘with a comparison of history teaching in both nations’ that details the political and contentious nature of history education (Mirkovic and Crawford, 2003: 91). Included in this analysis are topics such as the influence of state control over classroom knowledge, the selection process of textbook knowledge, and the relationships between interest groups involved in staking ‘their particular claim to defining the characteristics of nationhood’ (Mirkovic and Crawford, 2003: 48). However, a more complete sociopolitical analysis delving into the particularities of textbook production itself is lacking, as the authors do not enter a conversation around the politics of publishing houses and authorial selection. The bulk of the study moves towards identifying how the discourse, content and usage in textbooks ‘of photographic evidence and pedagogical approaches’ perpetuate each country’s emerging political context through drawing on linkages to national identity (Mirkovic and Crawford, 2003: 91).

The Georg Eckert Institute’s (2015) international mapping of the Holocaust in educational materials is by far the most robust and comprehensive holistic analysis we encountered. The Institute surveyed 89 textbooks across 26 countries from 2000 onwards with a methodology focused on the ‘semantic (contextual and narrative) rendering of the event in socio-political contexts worldwide’, discourse and narrative structures, content and the ‘didactic approaches authors adopted’ (Georg Eckert Institute, 2015: 30–31). Visual analysis was further incorporated into the Institute’s discursive analysis by requiring ‘researchers to record the visual narrative conveyed by the selection and arrangement or sequence of images’ (Georg Eckert Institute, 2015: 30). Although broader sociopolitical discussions of textbook publication were avoided, the Institute’s analysis met all five of the criteria set out in our study.

Other studies that appear holistic in nature but that do not explicitly adhere to our five-pronged recommendation are Christine E. Sleeter and Carl A. Grant (1991) and Sean Carleton (2011). Sleeter and Grant (1991) analysed a large corpus of 47 textbooks, 14 of which were social studies textbooks used across the United States. Their study assessed each textbook in its entirety to parse apart the level of representation, and attention was paid to ‘inequalities based on race, sex, disability, and social class’ (Sleeter and Grant, 1991, 79). Their methodology is also extensive, with six aspects including a ‘picture analysis, anthology analysis, “people to study” analysis, language analysis, story-line analysis, and miscellaneous’ (Sleeter and Grant, 1991: 82). The extent to which didactic activities feature in the authors’ discourse and content analyses, or if the authors considered each textbook’s sociopolitical context is not apparent.

Carleton’s (2011) study of 34 history and social studies textbooks used in British Columbia, Canada, from 1920 to 1970, is also a good example of a holistic lens. While he does not assess the didactic portions of the books, he does assess the other holistic categories to explore the ways they justify settler colonialism to non-Indigenous students. Carleton (2011: 115) conducts discourse analysis of both the content and the visuals of his data set, and he showcases how Indigenous Peoples are portrayed as savage and violent, and settler colonialism as ‘gradually bringing civilization to indigenous communities’. The wide periodisation of the study necessitates attention to the sociopolitical and institutional context of curricular production. Carleton (2011) discusses the effects of the interwar period, the Second World War and the Cold War on the Canadian nation-building project and its provincial curricula, and then aptly segues into textbook analysis.

There are a few additional studies in our data set that addressed the wholeness of the text with attention to two or more of these forms of assessment. In this regard, some of the studies already noted for conducting visual and didactic analysis could well be classified as engaging in a holistic analysis, such as Wenzeler (2003) and Abdou (2018), among others. We found that those studies that could be classified as more holistic often tended to provide a more cursory application of different methods or examined far fewer textbooks relative to studies that focused only on content or discourse analysis. We also found, however, that perhaps because of the complexity of analysis, those studies that embarked on a more holistic approach were often clearer and more explicit about their methods than other studies. Greater detailed procedures for methods offer the welcome possibility for replication by other scholars. The overall lack of holistic analysis is in many respects understandable, given the complexities of each approach, especially for individual scholars, and the limitations of space in publications. That being said, digital technologies have made combining qualitative and quantitative methods much more feasible, and, consequently, holistic approaches potentially much more manageable.

Based on our findings from textbook studies, our research team has embarked on a rare and ambitious holistic analysis that examines two history textbooks – one for elementary and one for secondary schools – from each of Canada’s 13 provinces and territories. The two co-leads of the research team selected textbooks and two chapters per textbook based on a focus on Canadian history, attention to the project’s main themes of historical thinking and civic engagement, and a breadth of historical time periods across the data set. While we do not analyse each textbook in its entirety using all methods of analysis, different aspects of our research design require attention to the wholeness of the text, as well as to the front and back cover, additional front-end information and the index. Our study begins with an NVivo-assisted keyword search of each textbook for pre-identified keywords related to the main themes. This keyword search assists with a later manual analysis of content and discourse by providing the number of occurrences and coding for usage of terms. Next, researchers conduct a manual reading of the chapters for each textbook, which includes content analysis (that is, frequency and conceptions of main themes), critical discourse analysis (dominance and usage of key words, including linguistic codes), visual analysis (inventory of images, including type, text, positioning and purpose) and didactic analysis (inventory of activities, including pedagogical approach, skills and purpose). Finally, one researcher is conducting a sociopolitical context analysis that considers the authorship, publication period, publisher, approval process and more of each textbook. Classroom usage of these textbooks will be part of the team’s research, but it will emerge from survey and interview data from the larger project. For all methods, the study requires that researchers answer specific questions related to equity, diversity, inclusion and decolonisation. For example, for the visual analysis, we ask: Are historical power imbalances, supremacies and hierarchies made visible or invisible in the images? This study is ongoing and resource heavy, including a team of over 12 research assistants and professors engaged in the work over the course of more than a year.

We have taken up this ambitious research method because an assessment of history textbook studies makes clear that to isolate only one or two methods would miss the complex ways in which students are taught about the past. Our team is thus among the few scholars to take up Grever and Van der Vlies’s (2017) call for a holistic approach to textbook studies. They assert that a holistic approach is the only way to make visible the ‘organization of historical knowledge [that] sustains the perpetuation of national narratives in history textbooks’ (Grever and Van der Vlies, 2017: 287). Grever and Van der Vlies (2017), like Hogarth (2017), Repoussi and Tutiaux-Guillon (2010) and others, argue that the immense power of textbooks over how nations engage with their past and what that means for their political futures is undervalued. That power is unaccountable without a fuller examination of the complex and interrelated structures within and in relation to the textbook. We agree with Grever and Van der Vlies’s (2017) conclusion that a holistic analysis, from composition to comprehension, provides the tools that researchers require to understand and harness the power of textbooks for education. It is the future for textbook studies.

We have demonstrated that the call for the field of textbook studies to provide analysis from composition to comprehension has gone unrealised with respect to an explicit and holistic approach. Since the textbook represents a fruitful area for analysis that offers a multiplicity of insights regarding what is valued by the educational system, it is imperative that scholars do not become mired in studying written text alone and in isolation from other aspects of textbook creation (Klymenko, 2016; Podeh, 2000). This article is a renewed call for scholars in the vibrant field of textbook studies to be attentive to missing methods. We call for an expansion of methods when approaching textbook analysis. Many studies employing content or discursive analyses have focused on the processes of exclusion or inclusion based on sociocultural categories within textbooks. While content and discourse analyses are important, their disproportionate use, to the exclusion of additional methods, negates fuller discussion of the intricacies of textbook research supported by a holistic approach.

Departing from Grever and Van der Vlies’s (2017) and Eckhardt Fuchs’s (2011: 27) reviews of history textbook studies that identified the movement of textbook analysis ‘from traditional content analysis’ towards ‘paths with a view to methodological diversity’ and holistic analyses, we found that even many years after these studies, researchers still overwhelmingly deploy narrow and traditional analyses. Overwhelmingly, the field still lacks attention to methods and, more specifically, to a holistic approach that broadens the researcher’s capacity to dissect the positionality and perpetuation of historical knowledge(s). Addressing missing methods will enable textbook researchers to further contest the often-uncritical embrace of history textbooks. Moreover, a more holistic approach can challenge textbook researchers to consider how we might expand upon our understanding of the textbook as an artefact of our societies.