Problematisation as possibility: using Foucauldian discourse analysis to study digital educational technologies

Abstract

This paper proposes Foucauldian discourse analysis as a critical methodology for examining the digitalisation of education, drawing on empirical material from a Swedish educational context. Using Foucault’s concept of problematisation, the study examines how digital technologies are constructed discursively across educational policy documents, media texts and empirical data from interviews, surveys and classroom observations. The analysis identifies various discursive formations, such as professionalisation, individualisation, technological determinism, efficiency, modernity and democratisation, which shape the conditions for education in different ways. Key findings reveal that digital resources are often constructed as the individual responsibility of teachers; as tools for individualised learning; as market products that enhance efficiency; and as both modernising and potentially harmful forces. These constructions risk fostering isolation among teachers and students, instrumentalising educational processes and reinforcing commercial and technological imperatives at the expense of pedagogical and democratic values. The study demonstrates that Foucauldian discourse analysis can identify discourses in different practices, thereby highlighting the contingent and political nature of digitalisation in education. By problematising assumptions that are generally accepted as truth, this approach opens up possibilities for alternative ways of thinking and acting within educational practices, moving beyond dichotomised debates.

Introduction

This article proposes Foucauldian discourse analysis as a means of critically studying the digitalisation of education. The approach is demonstrated by presenting a study that draws on empirical material from a Swedish educational context. The presented study examines how digital educational technology is constructed and includes reflections on how these constructions may condition educational practices.

Digital technologies in educational contexts have in the later decades become a global issue (Williamson et al., 2023; UNESCO, 2024) as digital technologies in general, and more recently AI in particular, are frequently discussed in educational debates and highlighted in educational governance documents. Sweden is an interesting case in this respect. The country is known for being an early adopter of digital tools in education and has introduced them extensively in government documents over the decades. Recently, however, Sweden has adopted a much more negative stance (The Guardian, 2023; Ljungqvist and Sonesson, 2022).

During the pandemic years, the discussion became even more pronounced, connected to calls for a rapid increase in the pace of digitalisation of education (cf. Jandrić et al., 2020). Digitalisation of education is in this study broadly defined as changes in technology with the aim of improving educational processes (see Agélii Genlott, 2020).

The debate concerning the impact of digital technologies on education is notoriously polarised. Digitalisation is either idealised as good and necessary, or dismissed as harmful and overrated (Bayne, 2024; Selwyn, 2013). Proponents emphasise potential benefits such as personalised, self-paced learning and new collaborative methods, while critics raise concerns about excessive screen time, data privacy issues, impaired learning and the importance of human interaction and traditional teaching skills (see step 3, ‘Discursive constructions’, in the following analysis).

Digitalisation of education is also repeatedly presented as though it was an apolitical issue (Høydal and Haldar, 2022). However, the fact that both the pedagogical and the economic aspects of digital technologies in education practices are closely linked to the society of today, as well as to the aspirations for the future, points to its deeply political nature (Williamson et al., 2023). Therefore, critically reflective ways of investigating the various ways in which digital technologies influence contemporary educational practices are urgently needed.

This article aims to propose Foucauldian discourse analysis, problematisations, as a means of conducting such enquiries. Its main contribution is this proposal.

This approach encourages qualitative researchers to seek out differences, absences, and local contexts rather than similarities, presences, and universal contexts (Kaufmann, 2011). In other words, it encourages us to think differently about digital technologies in education (Foucault (1985/1990), as cited in Christiansen et al., 2021).

Problematisation is both the method of analysis and the object of analysis. With this kind of analysis, the problems that enable changes in practices and techniques are focussed and viewed in alternative ways. The object of analysis, in this case digital technologies in education, can then take shape as more coherent, but also more challenging (Koopman, 2013). Qualitative empirical research employing this approach focus on questions about power as relations and the structure of social conditions, found to be much needed in this field of research.

The study presented in this article consists of a problematisation of digital technologies for education, showing how these technologies are constructed in different discursive practices. It ends by discussing how this can work in terms of shaping the conditions of possibility for teaching and learning in educational practices. Such results are therefore expected to benefit anyone interested in understanding and contributing to the debate about digital technology in education. Different education stakeholders can use the results to inform their practice and advocate measures to counteract the risks identified.

The article begins with a presentation of critical research to contextualise the study within relevant research fields. Next, it presents Foucauldian problematizations in a theoretical and methodological section after which the empirical study and its basis are introduced. Three analytical stages are presented as follows: (1). Collecting statements; (2). Identifying relations between the statements in terms of discursive formations; and (3). Identifying discursive constructions of digital educational technologies within each formation. This is followed by a discussion of the results of the study in terms of how they may function as educational conditions, before the text concludes with some final remarks on the contribution.

Critical studies of education technology

Over the years, Foucault’s work has been used in many critical education studies of relevance for this study. This rich tradition includes studies that have used his work to analyse different aspects of education policy (Ball, 2016) and demonstrate how policy shapes the space in which educational practices can take place (Holloway et al., 2024). Furthermore, studies have explored the potential of a ‘Foucauldian education’ in which critique is central, and the contingency of power, truth, and subjectivity form the basis (Ball, 2019).

Foucault’s influence on the research field of education and digital technology specifically is less clear (Hope, 2015), but is reported to be growing rapidly (Lee, 2020). In several texts with a sociological focus on education, Neil Selwyn (2016a, 2016b) problematises digitalisation. He highlights how digital technologies used in schools are often referred to by using terms like resources, tools or instruments, all of which draw on discourses of practical usability and thus construct these technologies as objects of utility and effectiveness. Selwyn (2016a) further shows how schools are often described as outdated, while digital technology is presented with labels such as new and revolutionary. Other studies have addressed the belief that digital technologies and pedagogical approaches have the potential to fundamentally transform education, offering greater freedom and a more personalised equitable and effective learning approach (Grimaldi and Ball, 2021).

An emerging research field examines the use of surveillance technologies in education, drawing on Foucault’s (1977) theories of discipline, punishment and surveillance using the metaphor of the panopticon (Hope, 2015; Zuboff, 2019) or by looking at technologies such as visualisation and predictive analytics (Williamson, 2016).

Studies have also focussed on the question of how long digitalisation should be seen as something new and specific. The concept of the post-digital (Jandrić and Hayes, 2020) indicates that the process of digitalisation, the introduction of digital technologies in education and elsewhere, has been completed. Digital technology should then be viewed as an integrated part of educational and other practices today, rather than as something external and added.

Critical research challenging prevailing assumptions concerning the use and impact of digital technology has been presented as an important research agenda ‘in light of the unanticipated, sometimes harmful, consequences brought by the accelerated diffusion of digital technologies’ (Christiansen et al., 2021: 1). From these research fields I particularly bring the following interest into the study: how policy conditions education in different ways, the complexity of power as networks, the performativity of discourse and the possibility of viewing the present as post-digital.

Foucauldian problematisations

In the presented study, ways in which digital technology is discursively constructed is investigated in terms of problematisations. As indicated, a starting point for discourse analysis is that the way in which an activity is (re)presented is crucial to the way in which it takes shape; discourse is performative. In this perspective, available ways of understanding digital technologies and teaching make certain ways of working in education possible and desirable, while overshadowing or even excluding others. All activities are conditioned by this process. For example, the tendency to focus digital educational practices on narratives of efficiency, learning, and competence can hinder the realisation of education’s democratic mission. Similarly, if schools are considered outdated, efforts may be directed towards updating or even transforming them using digital technologies while other well-established methods may be overlooked or discarded.

Few other theorists have been as closely associated with the concept of discourse as Michel Foucault, although he explicitly resisted being charged with the formation of any method or tradition. He constantly changed his thinking over time and never formulated a discourse-analytic method or theory (cf. Åkerstrøm Andersen, 2003; Peters and Besley, 2013). Rather, especially his later work, can be understood as a set of unsystematically presented analytical strategies, always linked to a specific problem – problematisations (Howarth, 2000).

Throughout his oeuvre, Foucault often formulated methodological strategies as a kind of reflection after a completed study. For example, the archaeological method used in The Order of Things (1966), can be understood as having been developed only a few years later in The Archaeology of Knowledge and Discourse on Language (1972). History is used to critically discuss the present through problematisations, bringing light upon the given assumptions on which they are based and how they are produced. Problematisations thus focus on the conditions of possibility that shape any present situation in specific ways. This makes them suitable for understanding contingency – how reality is constructed in certain ways that make other ways unsuitable, unthinkable or impossible. To critically investigate this is one way of aiming to open for other ways of thinking (Khan and MacEachen, 2021).

Foucault’s work is often divided into four periods (sometimes two or three). The first period is considered to be Heidegger-inspired, the second, archaeological, is characterised by an interest in the formation of discourses, the third, genealogical, focuses on how knowledge/power relations construct objects and position subjects and bodies, while the fourth is considered to be ethically oriented, focussing on problematisations of the self-creation of subjects (Dreyfus and Rabinow, 1983; Peters and Besley, 2013).

The analysis argued for in this article draws mainly the second and the third period. Each offer alternatives to understanding the beliefs that prevail at a given time. Archaeology is about mapping the systems of thought, episteme (Foucault, 1966), that define beliefs about something at a particular time. Genealogy examines how such systems of thought change through an analysis of power as productive. Since analyses relatively rarely stop at mapping the ways of understanding available at a particular context but go on to examine how these ways of understanding change over time, archaeological and genealogical aspects often complement each other in analyses (see Fadyl et al., 2013). Still, archaeology and genealogy do differ in important ways (see Garrity, 2010).

The concept of discourse, most used in archaeology, concerns how science (Foucault, 1966) organises objects and subjects through systems of thought at a particular time. The point of genealogical analysis is to show that the systems of thought that archaeology presents are the result of complex twists and turns, rather than the effects of any rational, inevitable evolution.

In genealogy, analysis is complicated by the fact that power has no direction but works in netlike relations. Knowledge and power merge when discourses are emphasised as conditions of possibility that construct objects and position subjects. Genealogy also highlights how power operates through what Foucault terms as technologies of power that govern conduct and thereby practices, among the examples given are confessions or classification (cf. Behrent, 2013; Foucault, 1978/1980).

Problematisation is a kind of critique but not in the form of a verdict, but rather an analytic activity aimed at exploring how our everyday phenomena and practices are shaped as responses to problems to make it possible to think differently about them (Foucault (1985/1990), here in Christiansen et al., 2021), their ‘aim is to be critical without being judgemental’ (Koopman, 2013: 97). The phenomena and practices that are under inquiry as problematisations, are seen as contingent – as arising in processes, through complex relationships, and not the causal, logical outcomes of any distinct historical cause and effect relationship. Contingency is what makes systems of thought change. Problematisations require a conscious and critical eye to the political arguments that frame the phenomena under study or risks becoming bogged down in the very assumptions and systems of thought it is supposed to investigate (cf. Kendall and Wickham, 1999; Koopman and Matza, 2013). Koopman (2011: p. 4) expresses this beautifully:

Foucault denaturalizes and de-inevitabilizes. It is commonly recognized that there is an enormous critical potential harbored somewhere within these revelations of contingency.

Genealogy is about highlighting how systems of thought act as conditions of possibility, both enabling and limiting, for the problems on which the analysis focuses. Systems of thought are not seen as expressions of an underlying meaning or someone’s hidden intention or purpose; problematisations are not about uncovering or revealing any hidden ideologies. This distinguishes such analyses from some ideology-critical analyses based on critical theory seeking to expose the manipulative effects of power through language and other texts (Christiansen et al, 2021).

Through problematisation it becomes possible to see how problematic phenomena are not naturally given but produced in response to specific problems. The aspect of production highlights the contingent nature of problematisations. Because they are not given, they can be changed.

To sum up, Foucauldian problematisations offer concepts that work as tools for examining how social issues and phenomena are produced in discourse. This helps us see that our ways of thinking (e.g. about technologies in education) are not naturally given and to examine where they come from to consider alternative possibilities. Foucault indicated that his rich flora of concepts can be used as a toolbox from which to choose when conducting such analyses (Holloway et al., 2024; Hope, 2015) proposing his own use of Nietzsche and Heidegger as an example (Peters and Besley, 2013).

Problematisation of the digitalisation of education – The study

The study presented in this article consists of a problematisation of one specific social issue, digital technologies in education. The study draws on various previous presentations of methodology, such as Christiansen et al. (2021), Fadyl et al. (2013), Howarth (2000), Kendall and Wickham (1999) and Åkerstrøm Andersen (2003). As there is no established methodology with a capital M for Foucauldian discourse analysis, the methodologies need to be invented for each individual study, adapted to each specific problematisation. This could, of course, be discussed in relation to scientific quality criteria such as replicability and objectivity. Foucault explicitly prioritises heuristic quality, that the studies have something important to say, over methodological rigour as such (see Howarth, 2000). This also means that the studies themselves and the researchers involved become part of the production of the truth being studied; there is no external position from which such studies could be conducted (see Holloway et al., 2024).

The empirical basis of the study

Texts from different discursive practices are included in the analysis: educational policy documents, media texts, and extracts from two empirical studies involving observations and interviews with teachers, students and school leaders in Sweden during the period 2011–2023. Research has indicated how nation states are adapting trends and reforms in education from each other through transnational processes of policy borrowing and lending (Steiner-Khamsi, 2012). The borrowed policies are then adjusted nationally (and locally) to fit political, socio-economic, and cultural conditions (Ball and Junemann, 2012). This development adds to the international relevance of investigations of policy related issues from different national contexts. The timeframe was chosen because digital tools were highlighted in Swedish government documents in 2011, and the empirical studies used as data material in the study were also conducted within this timeframe.

Foucault collected rich materials of statements and called the collections archives. The archive of this study is much more restricted and includes statements gathered from teaching materials, teacher and student interviews and surveys as well as classroom observations, all from a study on five Swedish upper secondary schools. It also includes media texts and policy documents. Since discourse analysis does not focus on the consciousness of individual subjects as meaning-makers, the statements should not be regarded as the narratives or experiences of the individuals. It is not of interest who speaks or acts, instead the focus is on what meaning is available.

The media texts in the study were searched for using the Media Archive Retriever database (see https://www.retrievergroup.com/product-research). After experimenting with different search phrases, articles from the Swedish national press in the years 2017–2023 were collected using the search phrase: skol* AND utbildning AND digitalisering AND process* (school* AND education AND digitalisation AND process*) to focus on aspects of change. This search resulted in 170 articles, of which 98 were deemed suitable for the study after a review.

I also searched for studies of digitisation processes in business and industry to look for absences in the archive. To do this, I scanned the websites of three major national media for business and industry in Sweden: Näringslivet (The Business World), Dagens Industri (Industry Today) and the media business website Resumé, and collected 27 articles that dealt with aspects of digitalisation as a process of change and were therefore of interest for the study. The choice of discursive practices for which absences will be visible will, of course, determine which alternatives are feasible.

The policy documents used were: The curriculum for compulsory school, preschool classes and after-school programmes 2011, Lgr 11 (Skolverket, 2011) and the curriculum for compulsory school preschool class and school age educare, Lgr22 (Skolverket, 2024), those were downloaded. The interview texts, observations and survey results for the study were all taken from research material from two previous studies that I conducted together with different colleagues: Unos Uno and The possibilities of digital learning (in Swedish: Det digitala lärandets möjligheter), both led by Åke Grönlund. I will refer to them as UU and PODIL. In UU we followed the implementation process of one-computer-per-child in 20 primary and lower secondary schools (different levels) situated in 11 different municipalities, for 3 years (see Andersson et al., 2016; Wiklund and Andersson, 2018). In PODIL we followed the implementation process of digital textbooks, padlets and smart screens in five lower secondary schools in different parts of Sweden (years 7–9, most students aged 12–14 (see Grönlund et al., 2018)). In both projects, we carried out annual surveys, interviews with teachers, students and school leaders twice a year, and multiple classroom observations. The teacher interviews were individual, and the questions were open-ended (Patton, 2002: 342). Teachers were asked about their experiences with technological resources in their daily work, their opportunities and barriers, their skills and their need for more knowledge. Students were interviewed in groups of 4–6 students with questions of a similar focus to the teachers, and school leaders were interviewed about their experiences in managing the implementation processes. The open-ended nature of the interview questions makes them suitable for this type of analysis, which aims to explore what works as true in a particular practice, time and place (see Carnegie and Wilson, 2019).

The fact that we planned and asked the questions, chose the search terms etcetera, means that we as researchers were heavily involved in the production of the data. The data were thus co-produced by us as researchers and the participants in the studies, as discussed above. The picture presented is therefore one in which I, as the researcher, have been actively involved at all stages.

A wealth of material was analysed in the study, but for reasons of space only very small parts of this material are made visible in the presentation of the study in the form of illustrative quotations for the sake of transparency in the presentation of the analysis.

As already noted, the concept of archive is here used in a more restricted and concrete sense. This also applies to the concept discursive practice. In Foucault’s work, a discursive practice is the entire system of rules that govern what can and may be said in a particular time and society. Here I use the concept to clarify the more general (textual) regulatory context in which statements are retrieved. Examples of discursive practices are the media debate or the teacher’s talk. It is thus a delimited use of the term that refers to the practice in which the statement is retrieved, like how the concept of genre is often used.

The empirical study comprises three stages of analysis followed by a discussion presenting possible implications for education:

1. Assembling statements

2. Identify relations – discursive formations

3. Identify discursive constructions in the formations

Step 1: Statements

The first step in the analysis is a kind of mapping and concerns statements in terms of presence/absence. In Foucauldian discourse analysis, the statement (in French énoncé) is the smallest unit (Foucault, 1972). Within a given discourse, the statement constructs objects, subjects, relations and actions through delimiting determinations. A statement can be linguistic or multimodal, textual, it can also be a tone of voice, appearance, colours or a particular school architecture. Discourses are constructed through relationships between statements producing truth.

To study discourse and the production of truth, it is necessary to gather statements.

Making presence visible in analyses is about recognising the statements that take shape in the discursive practices under study. During the examination of the data, statements are collected that are considered relevant to the problematisation of the analysis.

The statements collected in terms of what is present, that is, actualised, in the discursive practice studied, will be visible in later steps of the analysis in terms of formations and constructions. Identifying absence is methodologically more difficult. Especially in contexts characterised by a high degree of taken-for-granted assumptions, it is difficult to see what is missing but could be present. One way is to look at previous studies with a similar and different focus, for example from other disciplines that can focus on something other than their own familiar focus. Another way is to look at other fields that are similar to the study in various ways, to see what is present there. Absences can indicate things that are presented as not allowed, or things that are not even considered possible, which are even harder to imagine. In this study, texts dealing with digitisation in business and industry were used to look for statements that appear there in order to find possibilities that are missing from the archive studied. The idea was that these are similar but not identical processes and therefor potentially useful as comparison.

The collected statements of presence in alternative discursive practices showed, among other things, that working environment is a theme quite frequently actualised there, as for example in Tidningen Näringslivet (The Business Newspaper):

New technologies have not only increased productivity but also improved the working environment for companies and employees in most industries (Dahlfors, 2023).

Responsibility and trust are in this discursive practice constructed as something that management and the profession should own which puts pressure on both parts, as here directed towards management, in Dagens Industri (Industry Today):

. . . the advance of digitalisation is stressful in itself. That’s why you have to create clear structures and make employees feel confident in the company’s purpose (Karlsson, 2017).

Here, technologies are frequently actualised as borderless, built into the ecosystems of the practices rather than external, add-ons, in the journal Resumé:

Technology without borders. And when high speed and reliability are combined with computer-aided analysis, in the form of machine learning, even the most difficult barriers can be broken down (Andersson, 2019).

Aspects such as digital identity and the mission of the organisation linked to its digital identity were identified within the discursive practices included as comparison. These are all aspects that have not been actualised in the discursive practices related to the digitalisation of education in this study and hence are treated as absence, possible alternatives.

Step 2. Discursive formations

Having collected rich archives of statements, the next step in the analysis was to identify groupings in relation to the problematisation of the study. Discursive formations are the principles that govern particular sets of textual, verbal or material actions. The principles create groupings, patterns in the relationships between statements that make them belong together.

Foucault did not make clear what principles of belonging might be involved, but in the logic of problematisations this is connected to the problems to which they are produced as solutions. Looking for groupings, therefore, means conducting a search for coherence in statements, for relationships that make the statements belong together in larger clusters according to the problems they are presented as solutions to. This belonging together is also characterised by contingency, which is why systems of thought can change over time according to social dislocations. In Foucault’s texts these are superordinate formations, and in analyses the term has often been used to refer to larger formations. One example is Edward Said’s (1978) classic analysis of Orientalism as a discursive formation.

This stage of analysis examines which statements occur together, interweave, build upon and support each other. It thus focuses on which statements are actualised within the same ordering context. Concretely, the many statements from step 1, where worked through in an analysis process aimed at identifying such patterns. By this procedure, six discursive formations were formed in the analysis, each evolving around a specific, organising problem area (see Table 1):

Table 1.

Discursive formations.

Discursive formation	Organising problem
1. Discourses of professionalism	Problems connected to the profession and professional role such as work load, time issues and professional development
2. Discourses of education	Problems around learning and teaching in schools
3. Discourses of technology	Problems of technology, function, usability and user friendliness
4. Discourses of modernity	Problems connected to contemporary conditions, modernisation, development issues
5. Economy discourses	Problems connected to costs and market issues
6. Discourses of value and ethics	Problems connected to values and ethical aspects, such as surveillance, democracy and control

Step 3. Discursive constructions

The third step concerned the discursive constructions; how objects take shape within the discursive formations. The formations where thus investigated as producers, with an interest in what and how they construct objects and subjects by naming, materialising and positioning them in relation to each other. By examining each of the six discursive formations presented above, the following discursive constructions of digital technology in education and learning were identified. The basis for this analysis is explained in each passage below and examples are provided. In this step of the analysis, technologies of power – practices through which power operates (see Foucault, 1978/1980) were additionally identified.

In discourses of professionalism, digital technology is constructed as (individual) responsibility

Digital resources are constructed as the professional, didactic responsibility of the individual teacher. In statements from teacher interviews and observations, as well as policy documents and debates, the teacher is positioned (and positions himself/herself) as responsible for making sure that digital technology is used, and used in good ways, in the classroom. Aspects of working environment is only occasionally actualised in the professional discourse. It then acts as a counter-force by linking teachers’ time and energy to accountability. Accountability is a technology directed at teachers, for example expressed in terms of what teachers should, in governing documents:

Teachers should organise and carry out their work in such a way as to enable pupils to use digital tools in a way that promotes their learning (Skolverket, 2024)

Accountability is also an expectation that teachers are directing at themselves in their efforts to be good (digitalised) professionals, even if it means that they have to carry a heavy workload. This was apparent in surveys from the project PODIL where an increasing percentage of the teachers (74% in 2018 and 80% in 2019) stated that they worked hard to increase their use of educational digital technologies and that this was stressful for them (yearly project surveys, PODIL). It was also expressed in interviews from the same project

. . . then you feel that you are using the padlets, working digitally, behaving in a way - even if it is not the best way. . . But then it’s not good anyway - BUT this would have been ok in the beginning but now we should have come further (Interview PODIL, language teacher).

Such expressions of high self-expectations and experiences of failure in relation to them where reoccurring in the data. Teachers repeatedly excused themselves in the interviews for not meeting the standards, not being ‘digitalised enough’.

In educational discourses, digital technologies are constructed as individualising

Digital resources are constructed as enabling individualised learning, adapted to the conditions of each individual student. This partly takes form as a virtue, as individualisation is constructed as desirable in educational practices by the steering documents: ‘Teachers shall take into account each individual’s needs, circumstances, experiences and mindset’ (Skolverket, 2024: 2.2). But this virtue view is countered by more negative aspects of individualisation where students are left alone, as in the following quote from the PODIL-data material:

When the pupil loses motivation, does not understand what to do, has been given a too difficult task, or tries to work but works in the wrong way and instead just sits and hangs around or does something else - then it rather becomes unproductive lonely work (data comment, researchers, PODIL).

In the PODIL project, teachers often expressed a certain complacency about students’ cooperation, communication and knowledge development:

There is no room for sharing. Many people sit alone in their room - some succeed, some don’t (Interview PODIL, civics teacher).

Similarly, in the observation diaries from the classroom observations in UU, individualised classrooms where students work with their different tasks the construction of digital activities as individual are apparent:

Students work individually on individual tasks for less than 40 minutes. Silent work. Everyone does different things - different subjects? Several wear headphones (Observation diary UU, students’ year 7).

The annual surveys from PODIL revealed that individual student work increased as the digital resources where implemented. There is a discursive struggle between the construction of individualisation as enabling and as disabling.

In technology discourses, digital technology is constructed as determinant and unnatural

Technological development is in the analysed media discourse constructed as a given force of nature. Schools are then positioned as forced to keep up with the constant change. The use of technology is also constructed using of concepts analogous to natural forces ‘sweeping in’.

The wave of digitalisation has swept over Swedish schools, eagerly cheered on by forward-thinking politicians (Mannerheim, 2020).

Digital resources are constructed as ‘something else’ in relation to traditional or analogue resources. Digital technology becomes a separate object, inevitably linked to development and the future. Handwriting is a skill that keeps appearing as such a traditional recourse against which digital technology is compared.

Teachers: It’s a betrayal not to teach our students to write by hand [. . .] (Mannerheim, 2020).

In recent years, there has been a turn in the debate, in which paper books and handwriting are constructed as desirable for education. Digitalisation is, by this discourse, constructed as a strong force that needs to be weakened in order for education to return to more traditional means. Here, the digitalisation of education is constructed as a dangerous experiment, as in the following quote from Dagens Nyheter, a major daily independent liberal newspaper, referring to Torkel Klingberg, a neuroscientist and in Sweden well-known media personality:

Politicians, school officials and many educational researchers often talk about the positive effects of digitalisation. But according to Torkel Klingberg, there is very little scientific support for the “gigantic experiment”, as he puts it, that is now to be implemented in Swedish schools (Lerner, 2019).

The general tone in Swedish political debate and media discourse has gone from welcoming digital technology in education to warning against it and constructing it as a threat to young people and to education. In this discourse, paper books and handwriting are constructed as more desirable for education than screens. Digitalisation is thus constructed as a strong force that needs to be weakened in order to return to more traditional educational means. Digitisation of education is here largely constructed as a dangerous experiment.

In economy discourses, digital technology is constructed as efficiency-enhancing and as market brands

In economy discourses, digital resources are constructed as streamlining, as enabling simpler routines, more fun and thus faster learning and more efficient administration. Digital resources, both hardware and software, are constructed as branded products with different identity-creating values in a commercial market. Adaptation to commercial ideals, such as entrepreneurism and market economy, functions as a technology of power, as in the curriculum:

In this way, education will enable students to develop digital skills and an entrepreneurial mindset. ( Skolverket, 2024 ).

In economic discourses in the analysed media, digital literacy is constructed as that which secures a prosperous future, as in Dagens Nyheter:

Digitalisation will determine Sweden’s future as a leading nation (Braunerhjelm, 2018).

In contrast to the technology discourse presented above, educational technologies are positioned here, in a national tabloid, as a solution to educational problems and the economy as an ideal that education fails to meet:

The solution to building a modern school exists, Fridolin [at that time minister of education]. It is digitalisation! That’s why we - business, companies, society, individuals - have digitised and streamlined our world, our routines, our results. But schools and the education of our schoolchildren have completely fallen behind (Boisen 2017).

Such constructions present educational technologies as possibilities – as long as the principles of the market are respected.

In discourses of modernity, digital technology is constructed as colonising but unnatural

Digital resources where for a longer period constructed as ‘modern’, as new, forward-looking learning resources, and the importance of everyone developing digital literacy is repeated as a mantra, constructing other kinds of resources as outdated and less important. However, as noted above, there has been an increasingly strong counter discourse in the Swedish media debate in recent years that portrays digital technology in education as unnatural and potentially dangerous. For example, this discourse emphasises handwriting and reading books as more natural than using a keyboard. In this context, naturalness and authenticity function as technologies of power.

Stop force-feeding schoolchildren with screens: Today, Swedish schools are governed by digital learning materials in many ways. To some extent this is a good thing, as society demands a certain level of technical competence. But it can be too much of a good thing (Jönsson, 2023).

Screens in schools were a failed experiment. There is something incredibly strange in how Sweden has handled this issue. Children spend much of their day on screens. They move less, read less and use pen and paper less, while books are increasingly absent. Yet society’s response has been to put even more screens in schools. [. . .] - We are doing the opposite: we are making the biggest investment in textbooks in modern times [. . .] We know from research that you get a deeper understanding when you read a book, compared to a screen (Edholm, 2022).

From an earlier strong focus on digital recourses as vital to the modern condition, not least in educational practices, a discursive rupture has emerged that increasingly positions screens as potentially harmful and an alternative constructed as educationally superior.

Value discourses construct digital technology as (un)democratising

The curriculum presents, among other things, the fundamental values that are to underlie education in Sweden. Digital resources are there constructed as democratising through the opportunities for communication and cooperative forms of work they enable, and through the construction of digital literacy as a civic virtue.

The mission of schools is to transmit and embed fundamental values and promote students’ learning, thereby preparing them to live and work in society. [. . .] Students should be able to orientate themselves and act in a complex reality with a high flow of information, increased digitalisation and rapid change. (Skolverket, 2024).

Digital tools are proposed as vehicles for building civic competencies vital for democracy:

All learners should be given the opportunity to develop their ability to use digital technologies. They should also be given the opportunity to develop a critical and responsible approach to digital technologies, to be able to recognise opportunities and understand risks, and to evaluate information. ( Skolverket, 2024 ).

But, on the other hand, warned about as hindering dialogue, in an example from the analysed media.

Pupils learn to form opinions, but not to listen [. . .] Mass communication, and most of all digitalisation and social media, has meant that we surrounded by a ’noise’ that drowns out even important voices (Enkvist, 2023).

A discursive struggle is evidently taking place.

Discursive constructions of digital technologies as conditions for education

Through the various stages of this analysis, several discursive constructions of digital technology and its place and use in education have been identified. I argue that such constructions condition education in different ways. Against this background, I move forward discussing some of the ways in which the constructions of digital technology constitute conditions for education.

Lonelyfication

Digital resources are constructed as the professional and didactic responsibility of individual teachers, leaving them alone to plan, implement and evaluate teaching with digital resources. Technologies of power work on them from within, through their ambition to align to the constructed ‘good, digitalised teacher’ no matter the workload. This is further reinforced by the absence of discourse related to problematising the working environment. Digital resources are also constructed as enabling personalised learning, adapted to each student’s individual needs. This creates conditions that carry the risk of teaching becoming solitary, with students left alone to navigate endless amounts of information on the web in their hyper-individualised learning process.

Othering

Digital resources are constructed as ‘the other’ in relation to traditional, analogue resources. This may be constructed positively, as digital resources being the better option, or negatively. This condition of possibility runs the risk of contributing to a polarised situation that works against digital technologies becoming an integral part of everyday didactic practice. The use or non-use of certain digital technologies then risks not becoming a didactic choice that teachers and students have to make in their everyday work. This is reinforced by the absence of discourses that construct digital technologies as part of the school ecosystem. Proposals from researchers of a post-digital condition may serve to propose an alternative.

Instrumentalisation

Digital resources are constructed as streamlining, as enabling simpler routines, more fun and therefore faster learning, and more efficient administration. This fits well with strong discourses about effective learning that embed the field of education even without the involvement of digital resources. As a result of these conditions of possibility, alternative teaching and learning processes that take time and require slowness, reflection and unpredictability risk being left out.

Commercialisation

Digital resources, both hardware and software, are constructed as branded products with different identity-creating brand names and associated values in a commercial marketplace. As a result, the power of large multinational companies risk setting the terms of school activities in a very tangible way. This suggests that educational technology companies are increasingly becoming a power to be reckoned with in terms of curriculum theory and school conditions.

Colonisation

Digital resources are constructed in a technologically deterministic way as the learning resources of the future, and the importance of everyone developing digital literacy is repeated as a mantra that simultaneously constructs other types of resources as obsolete and less important. As a result, digital resources risk colonising school activities and strongly conditioning other didactic choices and ways of working. As an alternative counterforce, proposals pointing at a post-digital situation can be imagined.

(De)democratisation

Digital resources are constructed as democratising through the communication and collaborative ways of working they enable. This acts as a potential alternative counterforce to individualisation, instrumentalisation and lonelyfication. At the same time, the ability to monitor students’ entire learning processes by technologies of surveillance poses a potential threat to students’ individual and collective freedom in learning.

Concluding remarks

The study has proposed problematisations, a version of Foucauldian discourse analysis by presenting an empirical study, highlighting how aspects of digitalisation as conditions for education could be investigated. The findings support the problematisations in the sociology of education that Selwyn (2013, 2016b) has presented, in terms of how digitalisation risks constructing education in ways contributing to a school with an increased focus on the power of market thinking, competition and individualism – or lonelyfication. Presented findings from studies about the common strong focus on efficiency and competence when discussing digital educational technologies are also supported by the study. The questioning of digitalisation as individualisation or personalisation presented by Grimaldi and Ball (2021) where further challenged by this result. Presented earlier research proposed that schools could be places where digital technology forms an ecosystem, where technology is not a revolutionary or threatening addition, but an integral part of the everyday life of the school (Jandrić and Hayes, 2020). This was identified as absent in the material studied.

The purpose of this article has been to demonstrate the usefulness of Foucauldian discourse analysis for the study of educational practices and to illustrate how such studies can take shape. One condition of possibility identified in the analysis was that certain ways of working in education were made possible by available ways of understanding digital technologies in education, overshadowing or even excluding others. Through problematisation, we could see how the analysed aspects are produced in response to specific problems, which highlights the contingent nature of problematizations. Since they are not inherent, they can be changed. The concept of technologies of power proved useful for understanding how this production of truth takes place in the different discursive practices studied.

Throughout the analysis it has become clear how the study has consistently delimited and concretised Foucault’s methodological approaches. This is inevitable, as his studies are concerned with long periods of time and entire traditions of thought, rather than limited phenomena such as the digitalisation of schools. The intention here has been to argue that the principles of these studies can be fruitful in a more limited study, using different types of texts and practices.

One further striking aspect of the conditions for education presented in the analysis is that they all tend to start with the agents involved and therefore inevitably separate (digital) technology, context and user. This is an important aspect positioning both students and teachers alone. An alternative possibility would be to see the activity or action of education as central and the ecology in which it takes place as involving all these agents. This would then introduce the idea contained in the post-digital, where this division between human and non-human agents is not taken as natural.

The results of the study also support the emphasis on critical curiosity that several previous studies in different ways have emphasised. There is a danger in the very dichotomisation of the debate, as it risks preventing both critical acuity and such creative curiosity.

In the introduction, I put forward as a rationale for using Foucauldian discourse analysis that it would allow for contingency as a more reflective way of framing the problem of digital educational technologies than the one-sided celebration or condemnation, as also supported by Christiansen et al. (2021: p. 21):

applying a problematization lens to the study of digital phenomena implies writing a critique of digital technology that has both emancipating and generative potentials: the lens aims to elucidate ‘the price paid’ for engaging with a particulary digital phenomon [sic] [. . .] inspire more researchers to practice forms of critique where digital technologies are not assumed inherently oppressive, corrupt, or unethical. . . [. . .] aid engagements with speculative and performative approaches to technology critique.

As I see the six conditions for education presented above, Lonelyfication, Othering, Instrumentalisation, Commercialisation, Colonisation and (De)Democratisation - open to such a broader view, where digital technologies in education are problematised for the critical aspects that they have introduced.

The focus on contingency also suggests the possibility of change, reminding us that the digitalisation of education is not a necessity given by nature, but something over which we can take power and think differently. To take such action, it is imperative that we are aware of the problematisations at work and the technologies of power involved in the conditions of possibility by the production of phenomena and actions, so that contingency is not mistaken for unchangeable nature. I argue that problematisations is useful for this purpose.