This chapter examines the historical context and evolution of teaching methods in Swedish teacher education. In doing so, it historicizes the ongoing debates about the quality of teacher education in Sweden and the perceived lack of instruction in teaching methods. Drawing on analysis of guidelines, government reports, recommendations, and curricula spanning from 1842 to the early 21st century, the chapter highlights changes in political reasoning about teaching, learning, and the organization of teacher education. The analysis reveals that teaching has been conceptualized in various ways throughout history. Early approaches viewed teaching as a technique for organizing and monitoring student progress. Later, teaching came to be seen as an art requiring not only good knowledge of the subject matter but also creativity and talent. With the subsequent "scientifi?ation" of teacher training and its gradual incorporation into higher education in the late 1960s, the course on teaching methodology came to be perceived as normative, emphasizing reproducible skills and best practices. As a result, by the end of the 20th century, it was replaced by the course in Didaktik, which, instead of providing teacher students with ready-made methods, focuses on critical evaluation of teaching based on relevant research. This does not mean that the question of teaching methods has disappeared from Swedish teacher education programs. However, in the highly differentiated educational system, it is handled differently by different institutions.

The concept of differentiation holds immense significance in education, touching upon aspects like access, inclusion, justice, and equality. However, it is also a complex and elusive notion, which acquires different meanings across historical and cultural contexts. This article explores the shifting reasoning about differentiation in the Swedish educational context. Inspired by Foucault’s account of disciplinary power, it conceptualizes differentiation as a technique for marking and addressing gaps between individuals. Drawing on an analysis of governmental and scholarly reports from 1842 to the late 1960s, the article identifies three shifts in the reasoning on differentiation: 1) from differentiation by socioeconomic class as a given factor to the search for scientific rationales for differentiation based on measurement of intellectual ability, 2) from viewing differences in intelligence as biologically conditioned and stable to viewing them as amenable to training and correction through education, and 3) from a focus on inputs to a focus on outputs. Overall, the article argues that even if the term ‘differentiation’ itself has been discursively replaced by others, the ideas underlying it—the search for gaps—continue to shape education in Sweden and beyond.

Reading is said to have a unique place in human civilization, transcending times and cultures. It is considered crucial for school performance and the development of democratic citizens (Skolverket, 2022). However, according to the Programme for International Reading Literacy Study (Skolverket, 2023) reading habits in Sweden are declining. Overall, the act of reading is perceived to be at risk – a concern that has been eloquently articulated within and beyond the Nordic context. Declining reading habits are seen as a major challenge in contemporary societies, with potential negative effects on children’s cognitive, emotional, and personal development as well as on the formation of a well-rounded, critically thinking, and informed citizens.   In examining how reading is produced as a societal, scientific and political concern (cf. Latour, 2004a), we have elsewhere introduced the concept of the ‘reading industrial complex’ (Sundström Sjödin, et al., in press). This concept posits reading as a multifaceted matter involving a wide range of actors, each holding their own view of what reading is, what it entails, and why we should care about it. The present study focuses more narrowly on the role of science in shaping perceptions of reading as a valued activity, influencing policies, and informing pedagogical practices.

Theoretically, the study is inspired by Latour’s concepts of matters of facts and matters of concern (Latour, 2004a, 2004b, 2014) to explore how scientific knowledge about reading is constructed and transformed into established ‘facts’ or ‘concerns’. This involves uncovering the desires and aspirations behind research initiatives and examining the ‘laboratory life’ (Latour et al., 2013) of reading science.  To achieve this aim, we trace the shifting epistemologies of reading research as reflected in the Journal of Literacy Research from 1969 to 2022. For that, we selected over 200 articles which we coded in accordance with our analytical interest and the purpose of the study. Of particular interest was to explore how the value of reading is constructed within selected articles.   Preliminary findings indicate significant shifts in reading research over time. These include expanded conceptions of what counts as reading and literacy, increased interest in contexts and the use of qualitative research methods. Nevertheless, reading continues to be treated as essentially one thing, albeit complex and multifaceted, that can be observed, measured, and assessed, often at the level of the individual. In Latourian terms, then, we can say that while reading is widely recognized as a political and societal concern, it still tends to be approached as a matter of fact.      

This study is part of a larger project called The Fiction of Numbers, in which we locate and explore the intersections between the spheres of science, public discourse, policymaking and educational practices. We specifically examine how reading becomes a specific node, or discourse, where the changing ideas on societal, sociotechnical and educational imaginaries (cf Jasanoff, 2015; Rahm, 2019; Sundström Sjödin, 2017; 2019) and solutions take place. Drawing on Science and Technology Studies (STS), we are primarily concerned with how knowledge and facts are produced and naturalized; that is, how a phenomenon is produced as a matter of course and thus becomes difficult to question, and the ways in which values and politics of knowledge become invisible in this process (Dussauge et al. 2015; Latour, 1987, 1993).

In this sub-study, we “trace” – in Latour’s (2007) sense of the word – specific ‘numbers’ related to reading that are regularly referred to in media as well as in educational and political settings in contemporary Sweden. The numbers are used in reading promoting arguments: it is claimed that seventeen-year-olds who read a lot have a vocabulary of 50,000 words, while their low-reading peers have only 15,000 words in their vocabulary. It is also argued that 50,000 words is what is needed to be able to read and understand a typical newspaper text.

These kinds of numerical claims circulate in the public discourse and are often unchallenged and presented as matters of facts. Uncontested, the numbers are left to do their work – efficient in establishing truths, suggesting impartiality and transparency, ‘strengthened by the historical relationship between numbers and rationality, objectivity and control’ (de Wilde & Franssen, 2016, p. 505; see Hacking, 1990; Porter, 1995). They stabilize beliefs about reading into hard facts. By that, they also naturalize reading as something inherently good and useful, and therefore difficult to question (Sundström Sjödin, 2019). Moreover, although the construction of the problem implies the construction of the recipient, i.e. the so-called troublesome subject, in this case it remains unclear for whom exactly the lack of reading is a problem (Marres, 2005).

In this study, we trace the specific numbers we encountered in various sites of what we call “the reading industrial complex” (Sundström Sjödin et al, in press). We trace the origins of these numbers, how they have been produced, and with what tools. In doing so, we aim to develop knowledge on how reading is constructed as a public problem and a societal desire and what role numbers play in this construction. This aim is specified in the following three research questions: I) Which actors are involved in the dissemination of particular numbers related to reading, and who are the (implied) addressees of these numbers? II) What societal and educational imaginaries and desires are embedded in these numbers? III) What are the “origins” of the numbers? How and for what purposes were they produced? Theoretically and methodologically, the study draws on concepts and sensibilities of STS to explore the processes of knowledge production and dissemination, developed in the section below.

 What image comes to your mind when you hear ‘Blooms Taxonomy’? Most likely it is a pyramid with several different colored levels of knowledge from ‘remember’ to ‘create’, with implied or explicit arrows pointing upward. In fact, this visualization of taxonomy is one of the most popular. Yet, its origin remains a mystery: it was not part of Bloom’s et al (1956) original framework or the later revision (Anderson & Krathwohl, 2001). On the one hand, pyramids and triangles are a common way of visualizing theoretical models in the social and educational sciences: think of the didactic triangle, Maslow’s hierarchy of needs (1943), or Dale’s cone of experience (1946). However, while these models have largely retained their original pyramidal representations over time, Bloom’s taxonomy has evolved into various visual metaphors such as ladders, trees, circles, and flowers. What ideas about knowledge do these visualizations convey?

Developed in the 1950s, Bloom’s Taxonomy was designed to provide a wide range of educational professionals with a simple theoretical model that could be used to address curriculum and evaluation problems (Bloom et al, 1956, p. 1). Essentially a product of behaviorism, Bloom’s taxonomy emphasizes observable students’ behaviors resulting from instructions. Moreover, the very word “taxonomy” represents an attempt to apply models from the natural sciences, particularly biology, to the field of education. In biology, taxonomy refers to the classification of organisms into a hierarchical structure based on shared characteristics. By borrowing this concept from the natural sciences, Bloom’s Taxonomy sought to bring a similar order and ‘scientific’ rigor to educational objectives. A taxonomy, according to Bloom, unlike a simple classification system, must follow structural rules and reflect a “real” order among the phenomena it organizes (Bloom et al, 1956, p. 18). It is a method of ordering phenomena that should reveal their essential properties as well as significant relationships among them (p. 17). Recognizing the difference between classifying phenomena in the natural sciences and more abstract educational phenomena, Bloom noted that educational objectives, when expressed in behavioral terms, could indeed be observed, described, and thus classified.

Bloom’s Taxonomy has not only survived the decline of behaviorism but is still widely used in educational planning and evaluation in different parts of the world, including Europe (Anderson & Sosniak, 1994). Moreover, a new revision, known as Bloom’s Digital Taxonomy, was recently developed by Churches (2008) to account for the skills required in the digital age. Such persistence of the taxonomy can be attributed to several factors. First, its structured approach provides a practical and easy-to-use framework for educators and curriculum designers. Second, its adaptability to different visual metaphors may also contribute to its enduring appeal (see Mitchell, 2005). Third, most research on taxonomy tends to focus on its interpretations, misinterpretations and application in educational practice but ignores its historical origins, theoretical underpinnings, and visualizations.

This study explores the confluence of ideas and practices through which a hierarchy of knowledge is produced and disseminated as scientific facts. Specifically, it examines the assumptions and beliefs about knowledge implicit in the Bloom’s Taxonomy and its different visual representations. In doing so, the study brings together and extends the insights from a growing body of literature on how pictorial and graphic displays of conceptual models, methods or data transform ‘invisible’ phenomena into visible facts (Baigrie, 1996; Coopmans et al, 2014; Jones & Galison, 1998; Latour, 1993, 2017; Lynch, 1981; Pauwels, 2005; Rogers et al, 2021). This means that we regard pictures as an important part of discourses that establish ‘regimes of truth’ (Foucault, 2014) and promote certain ways of thinking, knowing, seeing, and acting in the world.

This article explores the complex interplay between the visual, numerical and verbal elements of data visualisation and their role in shaping policy concerns. Focusing on the aesthetic and temporal dimensions of statistical graphics and drawing on the notion of diagram in the Deleuzian sense, the article emphasises the performative nature of data visualisation. More specifically, it explores how data visualisation suggests, rather than reveals, particular visions of educational pasts, presents and futures. Based on an analysis of graphs and charts selected from recent UNESCO and OECD reports, the article discusses the practices of the datafication of time and temporalisation and the beautification of data, which together produce ‘beautiful evidence’. This evidence informs education policies and practices and affects the way education can be seen, known and acted upon.

Homework has traditionally been, and still is, a common practice in Swedish and Norwegian schools (Karlsson et al, 2019; Rogde et al, 2019; Westlund, 2004), serving as one of the key links between home and school (Borgonovi & Montt, 2012; Karlsson et al., 2019). Nevertheless, this practice is currently not regulated on a national level and is not even mentioned in curricula in neither Norway nor Sweden. Instead, every school, or even every teacher, has its own policy regarding homework.

In 2018, the OECD launched a pilot study titled International Early Learning and Child Well-being Study (IELS) which assesses the emergent literacy, numeracy, self-regulation, and social emotional skills of children at age five. These skills are described as fundamental for children’s future achievements in school and later on in adulthood (OECD, 2020). According to the OECD (2015), the IELS will eventually “provide information on the trajectory between early learning outcomes and those at age 15, as measured by PISA” (p. 103). Thus, the basic assumptions underlying the IELS is that intelligence and socioemotional skills can be objectively evaluated and compared, and that they are stable and predictable

Ironically dubbed the ‘Baby PISA’, the IELS has already drawn a great deal of criticism, which tends to be in line with that of PISA (Auld & Morris, 2019; Moss et al., 2016). However, despite obvious connections to other large-scale assessments, the IELS stands out in terms of its methodology which was developed for children who typically cannot yet read and write. The instructions were given by a pre-recorded voice on a tablet and children could indicate their preferred response by touching items or moving them around the screen (OECD, 2020). Thus, at its core the IELS relies on children’s ability to ‘read’ pictures and to match what they hear and see with what they know.

The use of visual imagery as a tool for measuring cognitive and socio-emotional development is by no means new. In fact, many intelligence and personality tests developed as early as the early 1900s (such as Binet-Simon intelligence scale or the Rorschach test) incorporated some form of images. Developed for diagnosing developmental or intellectual deficiencies in young children or to identify personality and mental health disorders, such tests provided a technique to reveal the invisible and to make the perceived differences between humans to become observable, measurable, comparable and, thus, ‘real’. Despite much criticism, tests of this kind are still widely used to differentiate ‘normal’ individuals from those ‘gifted’ or ‘at risk’ and to assign different pedagogical treatments to different groups of students (Paul, 2004).

By measuring the cognitive and emotional intelligence of preschoolers, the IELS marks the culmination of a century in which testing was of paramount importance. In this paper we situate the IELS within a broader history of image-based assessments to discuss how images function as a tool for differentiating students, controlling education, and predicting future risks (cf. Pettersson & Nordin, forthcoming). For that we trace the history of some of the most common intelligence and personality tests and outline the conditions of possibility that enabled image-based tests to appear scientific and to function as a source of evidence.

Data visualization has become an integral part of governing education, greatly expanding its reach and influence in the digital age. From tracking the performance of individual students to monitoring the overall success of educational systems, data visualization serves as a powerful tool for informing policy and decision-making on both global and local levels. By providing an easy-to-understand representation of numerical data, it helps governments to quickly identify patterns and trends over time, make calculations about the future and communicate complex information in ways that are both informative and aesthetically pleasing. While there has recently been an increased interest in understanding the role of numbers in shaping education policy (e.g., Pettersson, 2020), visual representations have so far received little attention. Given the importance attached to data in education governance (Williamson, 2016), this gap is surprising. The aim of this paper is, therefore, to contribute insights on how images, words and numbers work together to produce knowledge that makes educational systems amenable to analysis, comparison, and governance (Decuypere & Landri, 2021; Williamson, 2016). More precisely, we explore how quantitates are transformed into geometric shapes, arrows, bars, and vectors to create persuasive accounts of what ‘works’ and what needs to be fixed. We do so by analyzing abstract non-representational pictures employed by international education agencies (such as OECD and UNESCO) in their reports from the last three decades. Inspired by Science and Technologies Studies (Daston & Galison, 2007; Latour, 2012; Lynch & Woolgar, 1990), we consider data visualization a specific technique of knowledge production that structures our understanding of educational spaces and temporalities (cf. Decuypere & Simons, 2020). Although data visualization is often assigned the role of ‘cognitive aid’, the preliminary results of our study indicate that it is not as transparent and self-evident as it is widely believed. By allowing the viewer to ‘see’ the past and present and to imagine the future, graphs, charts, and diagrams convey the impression as if they were entirely devoid of politics. With this promise of objectivity visual representations turn invisible phenomena into ‘noisy’ but ‘beautiful’ evidence (Halpern, 2015; Lynch, 1991). Nevertheless, data visualization presupposes filtering of what can be seen, in what ways and for what purposes. As such, it operates as a mode of preemptive governance (cf. Massumi, 2007), whereby the visualized pasts and projected (un-)desirable futures are brought into and organize the present.