Bridging inquiry and critique: a neo-pragmatic perspective on the making of educational futures and the role of social research

ABSTRACT

The making of digital educational futures raises pressing social justice concerns. Against this background, scholars face the challenge of bridging the tasks of investigation and critical engagement. Inspired by French pragmatic sociology, this article presents the notion of plural school worlds as analytical anchor point for dealing with this double-task coherently and productively. Varying understandings of what makes ‘good and fair school education’ are identified as defining features of different school worlds. These understandings (1) are historically entangled with social and political orders, (2) inform how we envision, enact, and evaluate possible educational futures, and (3) define a shared discursive space for exchange between various groups of actors. We claim that the notion of school worlds thus furthers both our understanding of persistent patterns of disadvantaging in digital education and our capacity for critical dialogue with involved actors through sensitizing explication, shifting problematizations and reflecting performativity.

KEYWORDS:

• Social order
• educational futures
• critique
• school worlds
• French pragmatic sociology

1. Introduction

The digital transformation of education confronts social research with two kinds of problem: First, the empirical question of what kinds of educational futures are currently being instituted; second, the epistemological-cum-political question of how to contribute to the envisioning, enactment, and evaluation of these possible futures. The first question points to the task of inquiry: who is involved in the instituting of educational futures, on what grounds, and with what strategies and interests? What concrete configurations are emerging and with what consequences? The second question emphasizes the task of critical engagement: how to define one’s own role as social researcher vis-à-vis these processes?

The interlocking of the two tasks of inquiry and critique has gained considerable attention in recent years. After all, the digitization and datafication of education raises pressing social justice concerns (Dencik, Redden, and Treré 2019). These issues gain relevance with the raise of AI-enhanced educational technologies (AIED). Contrary to widespread hopes that AI-based algorithms will ensure equitable educational futures by radically personalizing learning trajectories and experiences (du Boulay et al. 2018; Steinberg 2023), these tools may actually reinforce rather than alleviate existing disadvantages due to their deep entanglement with relations of power and inequality (e.g., Dixon-Román et al. 2020; McStay 2020; Gulson, Sellar, and Webb 2022; Williamson 2017a; Williamson and Eynon 2020; Perrotta and Selwyn 2020; Dishon 2017). The promises of automation implied in AIED are inherently linked to risks of algorithmic discrimination, to reconfigurations of the pedagogical profession, and to questions of democratic deliberation and control (Smith and Fressoli 2021; Witzenberger and Gulson 2021).

Against this background, this article joins the quest for conceptual frameworks and tools that allow for both: guiding empirical research and reflecting about the ‘critical’ role we play as academics and researchers. We present the notion of school worlds (Derouet 1992; Leemann and Imdorf 2019) as a heuristic analytical device which accounts for pre-existing social, political, and moral orders while remaining sensitive to the situated agency and strategic capacities of various actors involved in the making and unmaking of digital educational futures. The notion of school worlds is inspired and informed by the Sociology of Critique (Boltanski and Thévenot 1999; Boltanski 2011; Boltanski, Honneth, and Celikates 2014), a strand of French pragmatic sociology (Guggenheim and Potthast 2012; Latour 2009). The Sociology of Critique starts from the observation that social actors can draw on a plurality of moral orders to coordinate in uncertain situations. In line with this suggestion, the notion of school worlds captures plural understandings and visions of ‘good and fair school education’ (Rahm 2021), i.e., of educational quality and justice. Thus understood, school worlds play a central role in the design and implementation of digital technologies and can therefore help unravel dynamics underlying unforeseen discriminatory potentials of educational technologies. At the same time, they define the very discursive space in which social actors (including scholars, teachers, and engineers) imagine, justify, or criticize educational futures – and hence provide promising anchor points for critical exchange and democratic deliberation.

In section 2, we discuss current perspectives on the bridging of inquiry and critique in the field of critical data studies and discuss how this double-task is linked with a fundamental ontological question: how to account for social order without falling back into trivializing understandings of causality and technological determinism (Facer and Selwyn 2021)? The issues at stake are most pronounced in the context of the ‘flat ontology’ of Actor-Network-Theory and related perspectives that play a dominant role in recent critical EdTech studies. In section 3, we introduce the neo-pragmatic notion of plural (school) worlds as an analytical tool that allows to conceive of social order without denying the critical capacities and strategic agency of social actors. On this basis, section 4 discusses how the notion of school worlds helps to grasp and define our own role as investigators and critics in relation to disputes over educational futures. To this end, we present some first conjectures that demonstrate how thinking in school worlds advances our understanding of how ‘novel interactions of convergent technologies, actors, and expertise are instituting socio-technical education futures’ (as the CfP to this special issue so succinctly states). Finally, we discuss how by using school worlds as anchor point we can develop strategies for bridging inquiry and critique by focusing on the tasks of sensitizing explication, shifting problematizations, and reflecting performativity.

2. The critical double-role of social inquiry: moving beyond flat ontology and descriptive complexity

Within just a few years, critical data studies in education have amassed an astonishing collection of insights that have changed our understanding of how educational technologies reconfigure learning and teaching. The originality of these insights is partly due to epistemological and methodological perspectives that have inspired the research field. These orientations are perhaps best captured in Actor-Network Theory (ANT), as it has been developed by Bruno Latour and others over the past decades (Latour 1996; Law 1992; Law and Hassard 1999). The innovative potential of ANT follows from three interlocked tenets: (1) a specific understanding of the role of critique, (2) flat ontology, and (3) descriptive complexity (Fenwick and Edwards 2010; Gorur et al. 2019). These tenets have been discussed in detail by other authors (Cresswell, Worth, and Sheikh 2010; Müller 2015; Couldry 2020). We therefore only briefly summarize them here: Concerning the role of critique, ANT rejects any form of assuming that the social sciences hold a privileged epistemic position for understanding the foundations and workings of the world. This epistemological position is combined with an ontological stance that Latour himself describes as ‘flat’ according to which ‘nothing should be reduced to anything else’. A flat ontology translates into a methodological orientation that starts ‘at the bottom’ to observe and describe how actors and objects form networks and assemblages that ‘hold together’, in ‘an attempt to radically free up our descriptive language so that we appreciate fully the actual plurality of the world’ (Couldry 2020, 1140).

In their combination, these three tenets offer a rich repertoire for grasping the complexity of technological assemblages by capturing their working in minute detail. In his nuanced reading, Perrotta (2021) summarizes the issues at stake by moving the problem of under-determination centre stage. He argues that the key analytical move afforded by ANT and related approaches lies in moving beyond naïve technical determinism, with important epistemological consequences (our understanding of causality) and political implications (our perception of objects and opportunities for political reflection, debate, and intervention). Appreciating the under-determination of any technological-educational assemblage regarding their actual effects allows both to account for the key role of actors’ critical capacities and strategic agency and to open spaces for democratic deliberation (Callon, Lascoumes, and Barthe 2009).

But while ‘[d]escribing how things occur is straightforward using ANT’, explaining ‘why things occur poses a challenge’ (Cresswell, Worth, and Sheikh 2010, 6). In John Law’s (1999) words, for ANT there is no social order, but only endless attempts at ordering. ANT invites us to tell ‘tales about how the world cannot stop transforming’ (Cresswell, Worth, and Sheikh 2010) – but makes it hard to tell a story about why so many patterns in the social world tend to persist. ANT empowers us to describe how micro actors become macro actors by growing and stabilizing networks (which we may read as the establishing of ‘social order’) (Callon and Latour 1981). However, we have little at hand to make sense of the existence of any form of ‘social order’ – recurrent relations and patterns that prevail across time and situations, such as durable patterns of educational inequality. The problem for ANT is that as soon as we presuppose some form of social order and aim to account for its persistent patterns and consequences, our ontology is no longer completely flat and our methodology no longer purely descriptive. Crucially, our understanding of the role of critique also hinges on how we conceive of social order, since critical interventions are inherently interrelated with existing relations of power and inequality (Boltanski 2011).

The situation becomes even more complicated if we appreciate that orders and patterns persist although they run counter to the stated intentions of involved actors and despite both increasing awareness and extensive knowledge about the dynamics involved. Macgilchrist (2019) discusses this kind of configuration as example of how the EdTech scene is troubled by ‘cruel optimism’, ‘when the object of desire is blocking one’s flourishing’ (77). For example, the strive to make educational inequalities visible through quantitative assessments and other forms of ‘datafication’ might in the end reinforce the very patterns one aims to abolish (e.g., by reductive aligning school education with the logics of standardized testing). The concept of cruel optimism is powerful because it demonstrates the intricate entanglements of existing orders of power and inequality, actors’ strategic agency, and the concrete effects and performativity of their situated practices. All the more, the question still remains: Why do actors not envision and enact technologies otherwise, once we know about them? How can we come to understand the agency of actors that – by definition – represent an already existing ‘social order’ (in the sense of pre-established powerful networks that already have a form and function), such as institutional and corporate actors? Why do narratives and projects by different actors look so similar, if all actors are involved in their own strategic network building? Finally, how can we assess the actual kind, depth, and degree of change taking place without some analytical and conceptual understanding of current ordered social realities in education (as in other fields)?

In order to address such questions, we require some conceptual framework that allows to deal with educational orders as part of wider social and political formations (Couldry 2020). In the following, we propose the notion of school worlds as promising starting point for this task. The strand of French pragmatic sociology that informs this move – the Sociology of Critique – has much in common with ANT, sharing among others a sensitivity for the role of objects, a methodological focus on situations of uncertainty and dispute, and a conception of social actors as capable of critical engagement and strategic agency. For the Sociology of Critique, however, the problem of social order including its innate asymmetrical relations of power remains centre stage, a circumstance that is mirrored in guiding analytical heuristics such as the notion of ‘worlds’ that we now turn to.

3. Introducing the heuristic of ‘school worlds'

In the following, we introduce the pragmatic understanding of ‘social worlds’ as a conceptual tool that allows to account for social order while remaining sensitive to actors’ critical capacities and strategic agency, thus allowing for novel takes on the relations between inquiry and critique.

The notion of worlds has been introduced and used by several scholars inspired by pragmatism, such as Anselm Strauss (1978) or Adele Clarke and Susan Leigh Star (2008). Howard S. Becker’s (1982; 2008) studies in the sociology of culture and arts provide a fitting example for the guiding ideas underlying a social worlds perspective. Becker employs the notion of ‘art worlds’ to capture the countless forms of cooperation and coordination involved in the production and evaluation of cultural products. Becker gives the example of lists of acknowledgements in movies to demonstrate the variety of tasks and responsibilities involved in producing art. Art worlds are stabilized by establishing expectations and dependencies between these various roles and contexts over time. They have no sharp edges and no fixed form. They develop and evolve.

There are numerous apparent overlaps between Becker’s notion of ‘worlds’ and ANT’s guiding concept of ‘networks’. Art worlds assemble actors and objects. The more interdependencies they can establish between the material world, organizational forms, and individual practices and biographies, the more stable they become. We can thus read Becker’s monograph as an illustration of how micro actors have become macro actors (Callon and Latour 1981) by forming durable and extensive relations between heterogeneous elements, resulting, for example, in some actors being classified as ‘artists’ while others remain in the category of ‘craftsmen’.

Becker however goes beyond descriptive accounts of networks of actors and objects by focusing on the semantic undergirding of art worlds as condition for their long-term existence. In his understanding, art worlds need to be anchored in conventions (Becker 2008, 40–67) – (more or less arbitrary) rules that social actors need to know and follow in order to coordinate in uncertain situations (Lewis 1975).

This coupling of ‘worlds’ and ‘conventions’ is further developed in several strands of post-Bourdieusian French pragmatic sociology which have developed in close vicinity to and interplay with ANT (Barthe et al. 2013; Latour 2009; Guggenheim and Potthast 2012). Luc Boltanski is one of the most prominent names associated with this perspective which has been referred to as the Sociology of Conventions (Diaz-Bone 2017), the Sociology of Critical Capacities (Boltanski and Thévenot 1999), and, building on Boltanski’s later work, the Sociology of Critique (Boltanski 2011).

In these theory contexts, the notion of ‘worlds’ is used in a manner that is highly reminiscent of Becker’s understanding: worlds are shared frames for action that allow social actors to coordinate in uncertain situations (Storper and Salais 1997). As with Becker, these worlds entail close relations between actors, objects, and discourses (Diaz-Bone 2011). Also congruent with Becker, worlds are seen as anchored in conventions that define the rules that actors follow to coordinate (Batifoulier 2001). In contrast to Becker, however, the guiding idea in French pragmatic sociology is that actors always have a plurality of such conventions at their disposal. Not countless, but several – ‘more than one and less than many’ (Mol and Law 2002, 11) – different forms of defining and handling situations that are widely accepted as adequate, relevant, and legitimate.

This idea of plurality and its implications are most notably developed in Boltanski and Thévenot’s (2006) monograph on ‘orders of justification’. Orders of justification define moral grammars that actors rely on to decide questions of worth, justice, and relevance in situations of conflict and dispute. Boltanski and Thévenot identify six such orders that constitute different ‘worlds’, each of which comes with its own understanding of the common good, quality, and justice: civic world, domestic world, world of fame, inspired world, industrial world, and market world. Any situation can be defined and handled differently depending on the kind of world that involved actors believe it belongs to.

Inspired by Boltanski and Thévenot’s ‘orders of justification’, Jean-Louis Derouet (1992; 2000) offers an early application of the understanding of ‘worlds’ to the field of education. For Derouet, a ‘school world’ mirrors ‘historical traditions that have sedimented and constitute a common understanding of what makes fair and good education’, traditions ‘which are organized around several different models that offer coherent ways of arranging the various elements that together form an educational doctrine’ (1992, 81, our translation).

‘School worlds’ are thus comparable to other prominent concepts that aim to capture the ‘fixed’ social ordering of school education, such as the ‘grammar of school’ or the structuralist ‘school form’ (Dussel 2013; Hofstetter and Schneuwly 2013). In contrast to these other notions, the concept of ‘school worlds’ however, first, introduces a conception of plurality within commonality: we need to acknowledge that there are several different historically established forms of accomplishing the key tasks of school education. Second, the idea of ‘school worlds’ suggests that we choose moral orders as main analytical anchor and starting point: plural understandings of what makes good and fair school education.

These varying conceptions of educational quality and justice go hand in hand with different definitions of the nature of learning, valued learning content, the purposes of school education, didactic approaches, relations between teachers and students, adequate spatial and temporal arrangements of learning, and adequate test formats, among others. We can speak of a ‘school world’ in a strict sense if there is a set of principles of quality and justice that allows to constitute coherence between these different aspects. Derouet and others (cf. Leemann and Imdorf 2019), have identified several such ‘school worlds’ that are relevant today and mirror wider moral grammars (Boltanski and Thévenot 2006) and thus inform the politics and pedagogies of school education, including the making of digital education. For the purposes of this article, we focus on the following three:

• The industrial school world is organized around the maxim of efficiency; the purpose of school education is to prepare a skilled and differentiated labour force; the focus of learning is on the acquisition of standardized curricular knowledge and skills; standardization and quantification are seen as key for ensuring fairness, quality, and efficiency.

• The civic school world is centred on the idea of the common good as it takes form in modern nation-states and sees school education first as basis for developing future citizens capable of political participation and second for the selection and advancement of future state elites; equality of opportunities defines a guiding value, legal regulation and pedagogical professionality are seen as basis of quality and fairness.

• The inspired school world focuses on the idea of education and learning as self-regulated process of unfolding free and authentic subjectivity; the objective is to develop creativity and individuality, the process of learning is conceived of as the development of innate potentials; individualization marks an overriding criterion for assessing both quality and justice.

These school worlds have been developed and instituted over time and in concrete historical contexts. The underlying conceptions of good and fair school education are partly explicitly stated in pedagogical and political programmes (e.g., by thinkers such as Rousseau, Dewey, Montessori, or Skinner). However, many more actors and fields of practice have been involved in the making and instituting of school worlds. The list of individual and institutional actors who have historically participated in these processes include politicians, trade unions, parents, churches, journalists, educational sciences, and many more.

None of these school worlds has ever been implemented in a pure and isolated form. What we observe and expect in real-world contexts is that different school worlds are assembled in configurations and compromises, more or less tension-ridden combinations of different school worlds. These plurality-within-commonality configurations offer an anchor for empirical research into the current landscape of EdTech as well as for reflecting the critical engagement of social researchers in disputes over educational futures.

4. A school-worlds perspective on digital educational futures

How can the notion of school worlds help make sense of and engage in the instituting of educational futures? In a nutshell, we see its potential in providing a concrete answer to the question of what exactly critical scholars can contribute to democratic deliberation. This answer gives directions for empirical research by suggesting a cumulative set of research questions and defines concrete tasks that social research can and should fulfil. This critical role unfolds around three crucial tasks: (1) sensitizing involved actors for tensions and possible detrimental implications of school-world configurations, (2) expanding and shifting forms of problematization that inform how actors envision, enact, and evaluate educational futures in the making, and (3) furthering reflexivity, including of the performativity of social research itself.

In the following, we develop this general idea step by step. We aim to demonstrate how the notion of school worlds serves as an interfacing device that allows to enable and empower critical engagement by drawing disperse empirical findings together. School worlds in this sense define an analytical ‘centre of gravity’. This, in turn, implies that there is no single empirical strategy to follow or method to employ when investigating school world dynamics – the power of the notion lies in relating ethnographic observation, discourse analysis, interviews, surveys etc. Therefore (and in line with the conceptual objective of this article), the discussion in this section draws on findings from existing critical EdTech studies and our own research, rather than sketching a single empirical study.

4.1. Step 1: clarify the critical role of social research

French pragmatic sociology is fundamentally motivated by a discontent with key features of ‘critical sociology’ as imagined by Pierre Bourdieu (Boltanski 2011). Latour (2004) sees two insurmountable flaws in the traditional understanding of critique mirrored in Bourdieu’s programme: (1) the need to claim a privileged epistemic position and (2) the implicit necessity to reduce ‘one kind of entity to another’. For Latour, these flaws culminate in a conception of the role of critique that mistakenly centres on the tasks of unveiling (the ‘real’ interests or forces behind some phenomenon) or denouncing (e.g., the ignorance, the malevolence, or the incapacity of social actors).

French pragmatic sociology diverges radically from this understanding by emphasizing that social actors are themselves capable of critical engagement and epistemic agency (Boltanski and Thévenot 1999; Barthe et al. 2013). There is no deeper or hidden truth beyond or behind the desires and concerns that drive situational agency. For Latour, our job as ‘critical’ social researchers is hence not to debunk, but rather to describe (intricate complexities) and to assemble (viewpoints and aspects). This idea is powerfully captured in the conception of technical democracy that Callon, Lascoumes, and Barthe (2009) propose. They promote the idea of hybrid forums formed around ‘matters of concern’ (Latour 2004): bring as many viewpoints and forms of knowledge as possible together in order to negotiate solutions that hold together.

But what exactly can and should social research itself contribute to debates in such hybrid forums? Would dense description and constructionist accounts really be appreciated as important contributions? Or might social actors be interested in more than just having their own complex positionalities adequately represented in sociological descriptions, as Gad and Brunn Jensen ask (2010)? A more active role would mean not to restrict social research to assemble (other social actors), but to move from being spectators to becoming spect-actors, in Lury’s (2021) powerful play of words.

A social worlds perspective provides a specific answer to what it is that we as scholars might actively contribute to debates with and between already capable and knowledgeable actors: One of our main tasks is to understand how actors move within and between plural social worlds in ways that they can in principle reflect and talk about but which they are seldom fully aware of. On this basis, we can identify unseen tensions between school-world references and discuss possible implications that actions taken on the basis of one school-world might develop in situations that follow another. By thus pinpointing contradictions and possible unwanted effects, we can contribute to the expansion and shifting of discursive orders, opening avenues for more (and more critical) engagement by involved actors.

4.2. Step 2: explicate school-world logics in accounts of educational futures

Following this understanding of the critical role of social research, the first task is to explicate school-world references. Actors involved in envisioning and enacting educational futures may be capable of navigating different school worlds, but they are not necessarily aware of that and how they refer to them. What is more, for actors involved in the development and implementation of educational technologies there is no way around anchoring their ideas in some established understanding of educational quality and justice that allows to define a valuable role for their EdTech in everyday learning contexts. This need holds for any kind of product, be it intelligent tutoring systems which train students to perform in industrial school worlds of standardized testing (Perrotta and Selwyn 2020), face recognition tools that allow to monitor student behaviour in classroom situations (Selwyn, Campbell, and Andrejevic 2022; Gulson, Sellar, and Webb 2022), the detection of emotional states in learning processes (McStay 2020), or classroom management dashboards (; Manolev, Sullivan, and Slee 2019; Jarke and Macgilchrist 2021).

This strong requirement to relate to established school worlds might not be surprising in established pedagogical and political arenas. But it also holds in social settings that on the surface follow very different logics, such as those depicted in current research on platform economics in education (see e.g., Decuypere, Grimaldi, and Landri 2021; Kerssens and van Dijck 2021; Komljenovic 2021). To give but one example from our own research, striking examples for the need to relate to established school worlds can be found in ethnographic observations at EdTech summits. Some of these events are organized to specifically train start-ups and young entrepreneurs and to bring them into exchange with possible investors. They are driven by market-oriented logics of maximizing return on investment. Only few participants at these summits have a background in education; schools, and students are primarily of interest as putative costumers. Accordingly, pedagogic reasoning seldom plays a role in these arenas. Nonetheless, start-ups pitching their products and investors making investment decisions need to relate to school worlds at several crucial points: when framing a problem that some technological product provides a solution for, when assessing the compatibility of their products with the real-world markets they are targeting, or when thinking about how to persuade teachers, principals, parents, and students to try or evem buy their product (Selwyn 2022). References to ‘good and fair school education’ may be more or less subtle or explicit, they may be more or less coherent, but there is no way around them because they define crucial interfaces between those who design and sell EdTech products and their imagined users and costumers.

The task of explicating school world logics is especially pressing in the context of AI-based automation (Beer 2017; Al-Amoudi and Latsis 2019; Gulson, Sellar, and Webb 2022). Assumptions on the nature and purposes of learning and school education are built into algorithmic black-boxes but remain opaque by design (Dourish 2016; Dixon-Román 2016; Kitchin 2017; Neyland and Möllers 2017; Perrotta and Selwyn 2020; Jarke and Macgilchrist 2021). The notion of school worlds provides a vocabulary for naming these moral orders. We need to ask and investigate: What moral, political, and pedagogical presuppositions are coded and hardwired into such algorithmic devices?

4.3. Step 3: explore how school world logics translate into situated practices

The notion of school worlds warns us against prematurely assuming that we always already understand how some technology will actually change educational situations. Many other actors and factors are involved in the implementation (or non-implementation) of technologies in everyday classrooms. With its focus on the interrelations between underlying moral orders (of quality and justice) and arrangements of objects, actors, and problems in everyday situations, the notion of school world can advance our understanding of how envisioned technological futures materialize in already ordered, yet uncertain situations.

For example, Ideland (2021) gives an impressive account of how EdTech assemblages go hand in hand with a reconfiguration of the role of the teacher and understandings of professional identities and roles. In turn, teachers are important mediators who structure the actual usage and effects of digital technologies by implementing them in concrete learning environments, in line with their own purposes and understandings. In our own research, we observed how patterns of adoption and adaptation can vary between teachers. Teachers with less affinity to technologies tend to strip technologies of features that do not seem useful for how they usually design learning environments. Thus, digital tools such as tablets can be easily re-interpreted into panoptic disciplining devices in ways hardly foreseen by any software engineer (Steinberg 2021). Other teachers see themselves more as ‘educreators’, personally adopting and incorporating the visions of digital educational futures. These teachers, typically, do not only or primarily adopt the technology but rather show strong tendencies to embrace the surrounding inspired-school-world-narrative and create learning environments accordingly. Making such (plural) school worlds conceptions explicit opens avenues for dispute and can thus enable and empower different actors (students, teachers, parents …) to become more critical and creative when dealing with digital learning tools. The explication of school worlds also warrants demands for more open dispute over conceptions of quality and justice and over the implications of imagined educational futures. The explication of school worlds also serves as a preparatory step for sensitizing actors for tensions and conflicts that may arise within school-world configurations, tensions that are often silenced in political and pedagogical notions of educational quality and justice (Horvath and Leemann 2021).

4.4. Step 4: identify tensions in school-world configurations

Once we have identified relevant school-world references, what tensions arise between countervailing moral orientations as well as between moral assumptions and the practical school world settings in which educational technologies and algorithmic black boxes are applied?

To give but one example, a school-worlds lens helps make sense of an often-noticed incoherence in the EdTech landscape. This tension has been noticed by various authors (see for example Dishon 2017; Chang 2019; Macgilchrist 2019; Sahlgren 2023; Rahm 2021). It unfolds between (a) a logic that is strongly aligned with behaviourist pedagogics (Knox, Williamson, and Bayne 2020) and standardized testing (Høvsgaard Maguire 2019; Thompson 2017) and (b) a logic that is more reminiscent of a pragmatic pedagogy and focuses on creativity, team working, and self-fulfilment (Dishon 2017; Rahm 2023). The notion of school worlds suggests reading this tension as mirroring a combination of two well established understandings of educational quality and justice: industrial and inspired school worlds. The inspired school world is most visible in programmatic statements, for example visions of how AI will change education. In our own research, we encountered impressive examples in the sub-genre of TED talks. Typical such TED accounts of AI in education criticize current school education for its stifling of talent and creativity and for the irrelevance of standardized curricula. School education cannot be considered good and fair because it hinders the free unfolding of innate talents and interests. This situation is seen as troubling due to the kind of skills which are needed in the digital era: problem-solving, team working, grit, and entrepreneurship. The educational project that follows from this problematization centres on the dissolution of established ‘boundaries of learning’ – age groups, school subjects, online and offline learning, and so on. The task of algorithmic devices is defined accordingly: to really get to know individual users and radically personalize learning content and learning trajectories across the lifespan and across learning contexts. Both standardization and efficiency – the key features of an industrial school world – are explicitly denounced in inspired school worlds (exactly because they are remnants of ‘industrialization’ – see e.g., the prominent TED talk by doyen of creative revolution Ken Robinson).

Such an industrial logic is, however, also strongly present in AI-related visions of education. It is most apparent in claims that AI will allow to overcome the alleged inefficiency of current school education. This other version of problematizating school education concentrates on the bureaucratic and organizational overload linked to teaching in classroom situations. The future role of AI, in this scenario, will be to automate tedious and tiring teaching tasks such as the grading of countless papers or classroom management (Jarke and Macgilchrist 2021; Gulson, Sellar, and Webb 2022). The traditional ‘boundaries of learning’ (Greenhow and Lewin 2016) remain very much in place in this scenario – algorithms will enhance efficiency, rather than disrupt traditional curricula, usual groupings of children, or the organization of learning content into school subjects. Thus, the conceptions of learning, of educational quality and of justice shift underhandedly, compared to narratives of how AI might yield a creative revolution of school education. Algorithmic devices are now meant to support the efficient acquisition of traditional curricula for all students with the aim of optimizing outcomes as they are measured in standardized achievement tests.

These tensions matter, since these two school worlds imply very different understandings of educational quality and justice, different understandings of teaching and learning, and different roles for algorithmic devices. School worlds are of course only ideal types and tensions between different understandings of quality and justice are what we expect. What we gain, however, is a more nuanced descriptive understanding of the complexity involved in the staging of educational projects – such as the designing and selling of EdTech products. This knowledge can be used to sensitize actors for these tensions and enable a more explicit dispute over the problems and objectives at stake. On this basis, other problematizations of educational futures and the (possible) role of technologies become possible.

4.5. Step 5: identify possible implications of tension-ridden school-world configurations

A school worlds perspective redirects the temporal focus of debate from imagined futures to the present and the past of school education. It underlines how actors strategically engage with existing educational orders to position themselves and their products as representatives of a supposedly inevitable future. Inspired and industrial school worlds play an important role for promoting EdTech exactly because they are already firmly in place. Taking these strong links to the present and the past into account allows to make sense of the striking regularities in current narratives surrounding educational technologies. The tension-ridden anchoring in the present and the past also helps understand the persistence of unwanted patterns despite the best intentions of involved actors, and despite increasing awareness and knowledge of involved risks (Macgilchrist 2019; Chang 2019). School worlds develop inertia and path dependencies and, for many actors, seem to define an indisputable reality.

On this basis, we can start thinking about how technologies become effective as complex sorting devices in concrete learning situations (Domina, Penner, and Penner 2017), potentially yielding unforeseen and unwanted discriminatory effects. Dixon-Román et al. (2020) provide an impressive example with their critical analysis of the racializing forces of seemingly neutral and harmless digital essay writing and assessment tools. Unwanted sorting patterns emerge on the basis of students’ learning biographies and their positional identities (Roth and Erstad 2016). These sorting dynamics take different forms in different school world contexts. For example, the standardized classifications that follow an industrial school world logic typically reproduce existing patterns of disadvantaging – they infer the likelihood of future failure from past disadvantages (Chang 2019; Perrotta and Selwyn 2020). In an inspired school world, other sorting dynamics are at play: in self-regulated and interest-centred learning environments, we will have to prepare for patterns of self-sorting on the basis of students’ self-images, biographically anchored interests, and perceived strengths and weaknesses (see the critical accounts of self-regulated learning in the digital era by Houlden and Veletsianos 2021 and Dishon 2017; 2021).

Such unwanted sorting effects mark a highly relevant entry point for critical scholarly engagement. The plurality of school worlds allows to decipher dynamics that result from the interplay of different understandings of fair and good school education, as well as to understand the possible effects that devices built according to one school world might develop in concrete learning environments that follow a different understanding.

4.6. Step 6: expand discursive spaces and shift problematizations

One key purpose of explicating school-world references and sensitizing for their tensions and possible unforeseen consequences lies in allowing for novel forms of problematization (Bacchi 2012; Lury 2021). By thus expanding discursive spaces, a school-world focus can help expand our room for democratic deliberation. As starting point for such a promotion of critical dialogue, we may start with asking how actors problematize current education to substantiate the educational futures they envision (see for example the findings presented by Ideland, Jobér, and Axelsson 2021). These envisioned futures need to be widely acceptable as relevant, feasible, and fair for other actors if they shall stand a chance to develop into serious projects of changing school education. They therefore have to be anchored in existing moral orders of education and in today's structural conditions of school education (including organizational, architectural, and practical requirements). In other words, they need to be anchored in some way in established school-worlds. Currently, dominant narratives on needed educational transformations are often reproduced without much hesitation. The power of these narratives lies in their effective ‘solutionist’ problematization of current school education (Ideland 2021; Ideland, Jobér, and Axelsson 2021; Morozov 2014). The instituting of alternative educational futures hinges on problematizing otherwise.

In the context of AI in education, such a re-problematization could be based on radicalized versions of a civic school world – which plays next to no role in current disputes over digital education (Chang 2019). In line with the idea of post-automation (Smith and Fressoli 2021), problematizing the lack of democratic control (missing accountability, transparency, monitoring, and regulation of algorithmic devices) could define a starting point for a re-centring of the debate on the risks and uses of educational technologies. This specific problem of lack of democratic control would, however, have to be framed in a wider narrative that also tackles the changing nature of democracy in an age of automation, the novel character of knowledge and information, and the need to convey the skills and the responsibilities to deal collectively with the multiple crises of our times (climate change, pandemics, resilient nationalism …). Crucially, we would have to underline that different conceptions (and measurements) of educational quality and justice are at play – which may lead to very different problematizations of current school education and understandings of the role of educational technologies should play in everyday school contexts (Chang 2019). The task of shifting and expanding problematizations further requires that we take the organizational conditions seriously that define how school education is currently imagined and evaluated. In other words, we need to take existing school worlds into account as inevitable starting point for any discussion of the future of education.

5. Conclusion

Our objective with this article was to introduce the notion of school worlds into the field of critical data studies in education. We argue that this concept offers a fruitful heuristic for deciphering the dynamics underlying disputes and projects concerned with educational futures. At the same time, a school-world perspective allows to draw direct links between our research and the ways in which social actors in various other fields of practice think and talk about educational futures, thus enabling the definition of a critical double role for social research that bridges inquiry and critique.

This article was concerned with the questions of how to engage proactively as critical scholars and of what contribution we as researchers can make. We would like to conclude with a brief note on the problem of performativity and reflexivity. Independent of whether we aim for such an engagement or not, we need to be aware that the social sciences have participated historically in the formation of school worlds (Derouet 1992) and they are always already involved in sustaining, evaluating, and reconfiguring them. The social sciences are inevitably interlocked with ‘the big Leviathan’, as Callon and Latour (1981) so vividly argue.

Social research necessarily refers to existing school worlds when framing research questions or forming arguments. We either affirm the relevance and acceptability of school-world arrangements by sharing their presuppositions or we challenge them by questioning their underlying assumptions. Either way, our research is intrinsically entangled with school world configurations. Boltanski (2011) describes this relation of social research to its surroundings as ‘complex insider position’. Scholars cannot escape their own performativity. It seems plausible to assume that this entanglement of social research with school worlds has increased over the past decades due to sustained demands for evidence-based practice and indicator-based forms of governance and accountability. By forcing us to make our conceptions of good and fair education explicit and to clarify how they relate to actors’ problem understandings, educational imaginaries, and everyday practices, a school-world perspective holds the promise of furthering our own reflexive capacities in crucial regards.

Disclosure statement

No potential conflict of interest was reported by the author(s).