Keywords

Introduction

In recent years, the production and use of numerical data have become increasingly important in curriculum policy-making. In a globalised and uncertain world, they provide a sense of predictability and calculability longed for by policymakers and practitioners (cf. Gorur, 2015). Numerical data offer an answer to the frequently asked question, ‘What works?’ They seemingly offer a secure place, free from any personal preferences or political ideologies, a place from which objective and disinterested decisions can be made about desirable futures based on quantitative measures. However, since formal education is a teleological practice, ‘a practice framed by a telos: an aim or purpose’ (Biesta, 2010, p. 500), the idea of making entirely objective and disinterested decisions about the future seems somewhat incongruous. The critical argument raised in this chapter is that curriculum policy-making based on numerical data is as normative as curriculum policy-making based on any other ideal. Data-driven curriculum policy-making here refers to a general trend among curriculum actors at all levels to increasingly rely on numerical data as a basis for their everyday decision-making (cf. Addey et al. 2017; Nordin, 2019). The different levels of curriculum policymaking referred to in this paper are classroom, curriculum, and societal (Deng & Luke, 2008), where the societal level, due to the transnational turn in curriculum policymaking, includes emerging transnational policy spaces and actors (cf. Anderson-Levitt, 2008; Nordin & Sundberg, 2018; Wahlström & Sundberg, 2017). The critique evolves in two steps in relation to the three levels of policy-making. The first step is deconstructive in character and draws on Porter’s (1995) work on the quantification of the modern democratic state and previous work on data-driven curriculum work in critically examining the constitutive ideas underpinning the discourse on data-driven curriculum work and how they are manifested at different levels of curriculum policy-making. The second step is reconstructive in character, making use of the non-affirmative theory in education (Benner, 2023; Uljens & Ylimaki, 2015, 2017; Uljens, 2016, 2018) to elaborate on a more reflexive position.

Data-Driven Curriculum Policy-Making

The role of and trust in numerical data as part of building the modern bureaucratic and democratic society has been thoroughly examined by Porter (1995), supplemented by science and technology studies (STS) focusing on the co-production of society and how the way we know and represent the world affects the way we choose to live together (cf. Jasanoff, 2004). Furthermore, sociologies of numbers (Woolgar, 1991) and quantification (Espeland & Stevens, 2008) have contributed to the understanding of ‘the relationship among science, numbers, and politics’ (Gorur, 2014, p. 353). In recent years, researchers have also started to relate these processes of societal quantification to education as a fundamental institution, producing and making use of numerical data for shaping societal futures (e.g. Gorur, 2014, 2015; Williamson, 2017; Lindblad et al., 2018; Mølstad & Pettersson, 2019). Special attention has been given to the role of international organisations (IOs) and the development of international large-scale assessments (ILSAs) in this process of shaping and reshaping the educational sector in general (e.g. Lawn, 2013; Lingard & Sellar, 2016; Meyer & Benavot, 2013; Steiner-Khamsi, 2003) and the way ILSA results are used in national politics for legitimation and/or delegitimation purposes (e.g. Addey et al., 2017; Alasuutari & Rasimus, 2009; Mølstad et al., 2017; Takayama, 2010). At the curricular level, empirical studies have been done on the implications of the quantification of society for the way schools are governed (e.g., Lewis, 2017; Lingard & Sellar, 2016) and official knowledge is organised in national curriculums (e.g. Mølstad & Karseth, 2016, Mausethagen et al., 2019; Prøitz & Nordin, 2020; Wahlström & Sundberg, 2017). Additional studies have also contributed knowledge on how processes of quantification reach classrooms, standardising and quantifying the everyday interaction among teachers and students (e.g. Au, 2011; Hardy, 2018).

Taken together, these strands of research have contributed widely to our understanding of how the production and use of numerical data have come to shape and reshape curriculum policy-making at different levels. However, still little research has been done focusing the normative ideas underpinning data-driven curriculum policy-making and how they (sometimes implicitly) operate as educational ideals within and between different policy levels. Furthermore, research on the production and use of numerical data in education still does not make use of educational theory to any great extent for a prospective analysis of what alternative educational imaginations are at hand.Footnote 1

In making use of the discursive and non-affirmative approach as developed by Uljens and Ylimaki (2015), the analysis in this chapter aims to contribute in both ways. It will critically examine some of the ways in which ideas of competitiveness, objectivity, and distance are discursively constructed and operating as educational ideals within and between different levels of curriculum policy-making, and it will make use of non-affirmative theory in order to elaborate on a more reflexive position from which alternative educational futures can be imagined.

A Discursive and Non-affirmative Approach

Discourse and Ideas

Making use of a discursive approach means acknowledging that education, above all, is a communicative practice whereby ideas about oneself, others, and society are shaped. Discourse, here understood in a non-idealistic way, is a social practice among others social practices constantly making and remaking the way people think and act. As such, discourse also includes a reflexive element when ‘people constantly generate representations of what they do as part of what they do’ (Chouliaraki & Fairclough, 1999, p. 25f). Put differently, people are constantly and simultaneously shaped by, and shaping, the discursive practices of which they are part. Transferred to the context of schooling, this means understanding formal education as composed of two interrelated dimensions, one that involves the reproduction of society and another that involves the transformation of that same society. In order to facilitate such a dialectical approach, Uljens and Ylimaki (2015, p. 39) talk about the formal curriculum as a ‘systematic interruption in the practitioner’s way of understanding herself and carrying out one’s professional tasks’. In this continuously ongoing process of discursive reproduction and transformation of society, Schmidt (2011) has pointed to the important role of ideas, and the substantive content of discourse. She argues that in order to understand institutional continuity (reproduction) and change (transformation), research has to take on a more dynamic view of both continuity and change by ‘concentrating on the substantive ideas developed by ‘sentient agents in discursive interaction that inform their policy-oriented actions, which in turn serve to alter (or maintain) institutions’ (Schmidt, 2011, p. 107). Questions of what is said have to be accompanied by questions of how and by whom it is said. The neo-institutional branch called discursive institutionalism developed by Schmidt (2008, 2010, 2011) distinguishes between three types of ideas, policy-, programmatic-, and philosophical ideas. Policy ideas, then, refer to a set of rapidly changing ideas at the surface level operating when ‘windows of opportunity’ (Kingdon, 1984) open. Programmatic ideas refer to more general sets of ideas when policy ideas are put into practice in institutional programmes. Such programmes are seen as the result of different actors competing and compromising on different ideas. Here, the role of expert knowledge is of great importance. These ideas are more stable than policy ideas, but still less stable than the third type, philosophical ideas, which refer to a deeper core of organising ideas, values, and moral principles. These are ideas that evolve over longer periods of time within the public, academic, and/or political spheres and sometimes operate implicitly as taken-for-granted public philosophies or shared worldviews. A discourse-oriented approach to curriculum policy-making that distinguishes between different ideas thus ‘provides a language for talking about the human interactive and interpersonal dimensions of any level of curriculum policy-making, from classroom to transnational’ (Uljens & Ylimaki, 2015, p. 39) and a critical lens to examine normative ideas operating within and between these different levels.

A Non-affirmative Approach

In addition to deconstructing prevailing discourses on data-driven curriculum policy-making, the critical argument made in this paper holds a reconstructive element. For that purpose, the non-affirmative theory (NAT) formulated by Benner (2023) and further developed by Uljens and Ylimaki (Uljens, 2016, 2018; Uljens & Ylimaki, 2015, 2017) is used. NAT positions itself theoretically as a reflexive approach and criticises both practice-oriented and critical-sociological approaches to curriculum policy-making for being overly normative in presenting ready-made futures. From the reflexive NAT position, practice-oriented approaches are seen as too instrumental, with conservatives missing out on education’s wider societal purposes, and critical-sociological approaches are seen as too prescriptive in defining in advance what values and norms should replace the existing ones (Uljens, 2018). Although NAT originated as ‘a theory in and for liberal democracy’ (Uljens, 2018, p. 11) and therefore itself could be seen as normative, it positions itself as more reflexive than the above-mentioned approaches. Instead of focusing on practical solutions or theoretical replacements decided in advance, NAT sees the role of curriculum policy-making as a means to reflect upon educational ideas and interests and their manifestation within and between different levels of education. Such a reflection could be seen as an invitation to open and democratic deliberation, recognising that different ideals, norms, and values exist but without affirming them. Instead of any content, the ideal or educational aim should be made subject to critical reflection by all parties involved, while simultaneously being open to renegotiating their own positions. This is not to say that affirmation does not occur at any time in educational policy or practice. Rather, it is to say that any kind of affirmation should be accompanied by an element of reflexivity, where the reflexivity, questioning what is taken for granted, is understood as the directional force rather than the affirmation of any given position (cf. Benner, 2023). Curriculum theory then becomes what Pinar (2011) refers to as a ‘complicated conversation’ to engage in rather than just a provider of ready-made solutions, be they practical or theoretical. This is not to deny the importance of educational research to provide solutions to practical problems, but in a time where the question of ‘what works’ echoes at all levels of curriculum policy-making, there is also a need for research that problematises, asks questions, and thereby causes problems.

We argue that in addition to enhancing the usefulness of educational research, that is, its capacity for solving problems, there is an ongoing need for research that identifies problems and, in that sense, causes problems. This kind of research challenges taken-for-granted assumptions about what is going on and what should be going on, and speaks back to expectations from policy and practice, not in order to deny such expectations but to engage in an ongoing debate about the legitimacy of such expectations—a debate that crucially should have a public quality and hence should take place in the public domain. (Biesta et al., 2019, p. 1)

To facilitate such an approach, NAT assumes relative independence (Uljens, 2016) between education and other social practices. Although acknowledging education as an arena of different actors with a variety of educational agendas and interests, be they economic, political, statistical, or other, NAT takes on a reflexive position relative to them all. Causing problems here means questioning any taken-for-granted assumption on what value, knowledge, or educational future is of most worth. Instead, any such ideal has to be found in and through communication that recognises all agendas and interests as recognised but does not decide on their directional value in advance. This is not to say that anything goes; instead, it means acknowledging a critical position of relative independence for educational actors at all policy levels.

In this paper, I will make use of three of NAT’s core educational concepts for identifying, and thereby causing, problems and challenging some of the normative ideas inherent in data-driven curriculum policy-making: openness, recognition, and summons to self-activity. Openness is here understood as both an invitation to communicative participation and as an approach to what is communicated. The openness to participation refers to communicative processes in which individuals develop their subjective uniqueness, as well as their cultural belonging, as pointed out by Uljens (2018, p. 20). In such an interactive practice, ‘personalisation and socialisation go hand in hand’. NAT thus acknowledges the dialectical relationship between the social and individual worlds and between the shared and subjective life worlds. But openness also refers to the participants’ approach to what is communicated, and here the concept of recognition is central. Recognition, with its roots in continental philosophy, on the one hand, recognises any participant as being radically free from the start. This is recognising the freedom of the other as a prerequisite for participation. In this sense, freedom involves the way participants are positioned and approached within any communicative practice. On the other hand, recognition also refers to recognising freedom as a result (of educational activities). Uljens (2018, p. 14) has talked about this as ‘The modern pedagogical paradox of freedom as a necessary assumption making education possible, and education as a necessary activity for making (cultural) freedom a possible consequence’. In this chapter, the use of the term freedom is limited to a general understanding of freedom as central to education, both as a prerequisite for establishing equal communicative relationships and as a result that needs to be actively maintained and developed within educational activities. The last concept used in this paper is a summons to self-activity, which refers to personal engagement in communicative educational practices. It is an invitation for equal participants to take part in communicative educational activities, which hold the potential for them to transcend their knowledge about themselves, others, and the world.

The non-affirmative approach in this paper is thus limited to the use of these concepts in order to identify problems and widen the educational imagination of what curriculum policy-making could entail.

Normative Ideas in Data-Driven Curriculum Policy-Making

In the following, I will give some examples of how the ideas of competitiveness, objectivity, and distance operate as educational ideals in relation to the societal, curriculum, and classroom levels of curriculum policy-making. The examples given below are chosen selectively in line with the purpose of the paper, which is to raise a critical argument on the normativity of data-driven curriculum policy-making.

Objective Data for Comparing Educational Progress at the Societal Level

Data and data infrastructures have become constitutive elements in governing school systems. Lawn (2013) has referred to this phenomenon as emergent ‘systemless systems’ built around the production and use of numerical data. As pointed out, this is a development to a large extent driven by international organisations (IOs) such as the Association for the Evaluation of Educational Achievements (IEA) and the Organisation for Economic Co-operation and Development (OECD), followed by others (e.g. Lawn, 2013; Lingard & Sellar, 2016; Meyer & Benavot, 2013; Nordin & Sundberg, 2014; Tröhler, 2014). In this section, I will point towards two fundamental ideas underpinning these ‘systemless systems’. The first one is the idea of competitiveness. Here, Tröhler (2014) has shown the importance of the crisis in the US caused by the launch of the Soviet satellite Sputnik in 1957. Although part of a complex historical web, this event is important in understanding an increased sense of transnational competition at the societal level that also affects education. Having lost the space race to the Soviet Union, politicians in the US called for the standardisation, centralisation and scientification of the educational system in order to reclaim international superiority. Tröhler (2014, p. 33) has described this as the Cold War becoming ‘thoroughly educationalised’. Education came to be understood as being built into an already existing conflict, positioning countries against one another on a competitive rationale. With the founding of the OECD in 1961,Footnote 2 the competitive rationale that emerged from Sputnik was recontextualised and translated within a neoliberal discourse focused on the role of education in developing competitive knowledge-based economies; thus still acknowledging education as a central means for success. Addey et al. (2017) reported on an interview with Andreas Schleicher, OECD Director for Education and Skills, in which Schleicher emphasised the importance for nations to measure the skills of their citizens since they are all drawn into a competition taking place on a global market. In order to relate to others in a competitive environment, one has to measure and rank. Schleicher and Zoido (2016, p. 383) put it this way:

In a globalised world, the yardsticks for public policy in education are no longer national goals or standards alone, but increasingly the performance of the most successful education systems internationally. The example of PISA shows that data can be more influential than administrative control or financial subsidy, through which we traditionally shape education policy and practice.

Significant for both the example of the Sputnik crisis and the rise of a global knowledge-based economy is that education is considered primarily as a means to another end outside itself, be it technological development or improved results on international large-scale assessments. As such, it does not have to answer for any educational ideals of its own, instead, it is the context that decides what is of educational value. However, I would argue that scientific objectivity operates as an educational ideal inherent in the practice of decontextualised comparisons of educational results. Only to the extent that data are made objective are they considered comparable. The idea of competing in education as a means of economic growth and prosperity, therefore, relies heavily on the possibility of producing objective and measurable outcomes.

From a non-affirmative position, a fundamental problem here lies in the one-sided focus on thinking of education in terms of comparisons visualised through ranking lists and league tables. Countries are positioned against each other as competitors in hierarchical relationships rather than as collaborators. Some countries are identified as role models, while others are made subject to extensive reforms in light of the best-performing countries in the world. As described by Schleicher and Zoido (2016, p. 375), ‘International comparisons can show what is possible in education, in terms of the quality, equity, and efficiency of the educational services achieved by the world’s top-performing educational systems…’. Thinking of what is possible in education thus boils down to the whole world looking at about a handful of countries. The imagination asked for by Schleicher and Zoido is therefore a narrow one, where ideas of competition and objectivity are taken for granted when calling upon the world to use its educational imagination. From a non-affirmative position, I would argue that this is quite the opposite of a widened imagination of what is possible in education; such a widened imagination can take place only where educational institutions and their actors are acknowledged a priori as relative independents relative to any other agenda, be it technological or economical. Positioning nations in a hierarchical relationship may enhance competition, but if extending imagination beyond such an approach is desired, a more radical openness has to be adopted. Following the non-affirmative rationale, the answer to what is of educational (and societal) value can never be decided in advance, instead, it is a question continuously seeking its answer in the ongoing complicated conversation (cf. Pinar, 2011). From a non-affirmative position, such a conversation transcends a scientific discourse occupied with categorial issues of right or wrong, true or false, or objective or subjective. The radicality of the non-affirmative education theory instead lies in its approach—any issue has to be made subject to profound communicative interaction among actors already positioned as free, while simultaneously searching for a widened freedom through open communication.

Standardised Classrooms at the Curriculum Level

Although to a lesser extent spelled out, the production and use of numerical data also frame what is possible to think and do at the curriculum level. As expressed by the OECD in the strategy document Education 2030, curriculum work, globally, is supposed to adhere to a transnational discourse in which education is seen as a means in a battle for global superiority and in which the production and use of scientific (statistical) knowledge have made this race possible.

In the face of an increasingly volatile, uncertain, complex, and ambiguous world, education can make the difference as to whether people embrace the challenges they are confronted with or whether they are defeated by them. And in an era characterised by a new explosion of scientific knowledge and a growing array of complex societal problems, it is appropriate that curricula continue to evolve, perhaps in radical ways. (OECD, 2018, p. 3)

However, as shown, by the radicality of curriculum development, framed by a strong normative discourse emphasising comparability between countries, municipalities, and schools and prescribing in advance what knowledge content is of most worth, the educational imagination instead runs the risk of being radically reduced (Au, 2011; Nordin, 2019; Sivesind, 2014). With reference to the US, Au (2011, p. 30) shows that ‘in the case of high-stakes testing in the US, as the content of the curriculum moves to match what the tests require, the structure of curricular content knowledge similarly shifts towards the fragmentation demanded by the test’. Although the OECD calls for radical curriculum development, it thus seems as if what is possible to imagine is predetermined by the ideas of competition and objectivity operating at the societal level and communicated in a variety of ranking lists and league tables. In a study of the 2011 curriculum reform for the Swedish compulsory school, Sundberg (2018) has identified a similar development. He shows how transnational as well as national political demands for comparability have led to an increased focus on developing a standards-based curriculum in Sweden with a close alignment between goals, content, results, and assessments. Among other things, the result also shows how this has led to a decreased local space for teachers to act as co-constructors of curriculum, positioning them as implementers of ready-made standards to be managed in the classroom. Furthermore, the results confirm what Au (2011) found in the case of the US; what is to be assessed determines what is being taught. ‘A normative pressure for the summative assessment of teachers and their teaching is prevalent in the wake of the standards-based curriculum. One teacher expressed it with candour: ‘Well, everything we do must be measurable’ (Sundberg, 2018, p. 127). Although touching briefly upon the results from the two studies, they seem to strengthen the hypothesis that a focus on objectivity and comparisons at the societal level also strongly frames what is possible at the curriculum level, where facilitating measurability holds an almost hegemonic position ‘reducing the richer, deeper meaningful learning into measurable outcomes’ (Sundberg, 2018, p. 129).

At a basic level, it seems as if openness, both as a prerequisite and a goal, becomes heavily reduced when framed within a discourse of measurability. As shown in the Swedish study, standardised classrooms also mean a reduction of teachers’ professional space, allowing for different forms of knowledge and experiences to inform their professional decision-making. From a non-affirmative position, standardising classrooms is problematic when done to make teaching and learning more effective and their outcomes decontextualised and comparable. Instead of promoting openness and qualified self-activity, it could be seen as an expression of what Au (2011) has talked about as a second wave of Taylorism.

The decontextualization, objectification, and subsequent quantification of students through standardised testing do not stand alone, however. These interrelated processes also make high-stakes, standardised testing an ideal tool for the New Taylorism because turning students (and, by extension, teachers and teaching) into decontextualized, numerical objects also frames students as products and places education firmly within the paradigm of factory production. (Au, 2011, 38)

Drawing on Au, the argument of the OECD (2018) in Education 2030 becomes problematic, advocating a one-size-fits-all solution for all countries as well as for every individual in order to not be defeated by the challenges posed to them by an increasingly complex and uncertain world. Offering one tool for every situation in a complex environment seems extremely risky. In such a position, opening up for a shared discussion on what alternative educational scenarios are possible and what different kinds of tools could be used for different purposes, then, offers a radically different approach. Adopting a non-affirmative position here could then be described as a form of communicative ‘destandardisation’, complicating the conversation and questioning presumptions instead of mechanistically standardising education in a factory-like manner.

Digitising Human Interaction at the Classroom Level

The ideals operating at the above-mentioned levels of curriculum policy-making also relate to the way teachers and students interact at the classroom level. Using Sweden as an empirical example, there has been an ongoing debate for decades about how to improve the equivalence of the Swedish school, a debate that is increasingly becoming part of the discourse of the standardised classroom, facilitating decontextualised comparisons. The need to equalise the selection, organisation, and assessment of subject content in order to minimise national differences has, during the last decade, led to initiatives such as a new curriculum explicitly aimed at minimising the space for teachers to interpret differently and the National Agency for Education producing supplementary material to help teachers plan, teach, and assess the curriculum in a more coherent way, initiatives taken to standardise every aspect of the classroom. A recent initiative in this pursuit of national equivalence is the experimentation with the digitisation of national tests (National Agency for Education, 2019), which were first tested in about 100 schools and are scheduled for full implementation by 2022. Without going deeper into the reform, it is of interest to mention since it strengthens the argument that human interaction is increasingly constructed as a problem within today’s discourse on data-driven curriculum policy-making. In order to come to terms with unreliable performance data, the evaluative process has to be cleansed of human interaction and contextual information that lead to biassed judgements. Distance between teachers and students thus operates as a normative idea underpinning this striving for equivalence (cf. Nordin, 2018). Furthermore, it reduces the way assessment is perceived as an integral part of teachers’ professional and specialised knowledge (e.g. Lasky, 2005). The pursuit of measurability redefines the role of teachers and students and their professional relationship, replacing closeness with distance, interest with disinterest, and reducing complex educational actors to assessable entities. Taking on a non-affirmative approach, such a reduction deprives teachers of their capability to critically examine the contextual circumstances leading up to the present position, and thereby their potential to transcend that very position. Instead of recognising them as culturally free, teachers (and students) are locked up in a narrow discourse with numbers as the only language available for communicating educational matters (cf. Nordin, 2018, 2019). However, although locked up in a narrow discourse, NAT reminds us that teachers, as relatively independent educational actors, always hold the power to make any taken-for-granted assumption the object of critical reflection with their colleagues and students. Engaging in an open and reflexive classroom discussion with other teachers and students on the substantive ideas constructing the discourse of data-driven curriculum work could then be seen as an act of taking professional responsibility, offering a communicative way out of mechanistically and unreflectively passing on prevailing ideas to the generations to come.

Concluding Remarks

The critical argument made in this chapter is that, despite claims of objectivity and disinterestedness, the constitutive ideas making up the discourse on data-driven curriculum policy-making themselves operate as ideals shaping and reshaping education. Organising curriculum policy-making along an evaluative rationale can therefore by no means be seen as a neutral act; it instead promotes a very specific and narrow perception of what education is or could be. In this paper, I have touched briefly on some of the ideas inherent in the discourse on data-driven curriculum policy-making in order to problematise the way they are operating as such ideals shape education today.

Educational Ideas in Data-Driven Curriculum Work

Recognising the ideas operating in data-driven curriculum policy-making as educational ideals means that they can be critically examined within the context of education theory. Such an examination facilitates a discussion of what is reduced and reshaped in data-driven curriculum policy-making and the exploration of alternative imaginaries is possible.

Tracing backwards the production and use of numerical data in education reveals a competitive rationale driving education towards producing objective and comparable outcomes. Education is treated as a means for competition, placing societies, regions, schools, teachers, and students against each other in a competitive relationship. Although international organisations and national governments try to repackage this rationale in terms of mutual learning and sharing of best practices, they do so within the context of competition, which, at the end of the day, results in different kinds of ranking lists and league tables claiming to be telling the truth about your country, your school, and yourself. Although seldom discussed, it seems equally possible that such a competitive approach to education can contribute to the fostering of protectionism and nationalism, sharing some knowledge and experiences while leaving out others in order not to lose competitive advantages. The idea of competitiveness operates as a philosophical idea on the societal level, coordinating the discourse about curriculum policy-making among transnational and national policy actors. As a result of this competitive approach to education, the production and use of objectives, and thereby comparable educational outcomes, have become imperative, not just to policy actors and politicians, but also to school administrators, teachers, students, and parents. In some respects, it could be argued that ranking lists and league tables promote an inclusive discourse on education since everyone can participate intuitively without any expert knowledge. Problems arise when such a discourse starts to permeate the mindset of educational actors as telling the whole truth about education, shrinking the educational imagination to questions of what works, and finding the most effective solutions to problems communicated as numbers in ranking lists. Another problem raised by Au (2011), among others, is that assessments never operate in a vacuum; they tend to direct what content is selected and organised in the curriculum and the way teachers teach in order to facilitate such a curriculum. In doing so, the idea of objectivity also operates as a normative ideal and an organising principle at the curriculum level. In the pursuit of comparability, objectivity itself thus turns normative, becoming an educational ideal. Using the digitisation of national tests in Sweden as an empirical example, I have furthermore argued that the idea of objectivity also contributes to reshaping the interaction between teachers and students in the classroom. When, for example, teachers contextually informed decision-making is replaced with digital procedures as a basis for assessment and an increased amount of supplementary material for teachers is produced in order to standardise classrooms, teachers’ ability to make professional judgements is questioned and their interaction with students reduced to what is prescribed in these standardised materials. An idea of distance thus operates at the classroom level in order to facilitate the objective outcome necessary for decontextualised comparisons (cf. Nordin, 2018).

Examining the discourse of data-driven curriculum policy-making thus reveals several normative ideas only touched upon here. However, acknowledging them as educational ideals opens up for a critical theoretical examination of the ways these ideals actually affect education as an end in itself and not just as a provider of solutions to problems external to education. If this is not done, they will continue to operate implicitly as integral parts of data-driven discourses, searching for ever more refined sets of objective numerical data as the obvious solution to any educational problem without ever having to answer to education theory.

Towards a More Reflexive Position

Examining the discourse of data-driven curriculum policy-making shows that there is no such thing as an objective spot from which disinterested decisions can be made. Curriculum policy-making is always framed by a telos. Characteristic for data-driven curriculum policy-making, it seems this telos is decided upon primarily by actors and organisations whose main interest lies outside education, be it economic, political or other. This might lead to a problematic situation for educational research if, in order to be perceived as relevant, it places too much focus on solving problems external to education, missing out on the critical examination of the ideas underpinning such problems. Following Biesta et al. (2019), I would argue that, under prevailing circumstances, educational research identifying, and thereby causing, problems is of equal importance, speaking back to any problem for education to solve.

In this paper, I have turned to NAT in an attempt to make a critical argument, speaking back to the discourse of data-driven curriculum policy-making. NAT offers a reflexive approach to any question of ‘what works?’ stating that neither problems nor solutions can be decided upon in advance, instead they should be made subject to intersubjective communication among participants recognised as equal and free. It is in such an interaction with others (and their otherness), that subjects can develop ‘a personal uniqueness and a cultural belonging, i.e. personalisation and socialisation go hand in hand’ (Uljens, 2018, p. 20), and problems find their solutions. Furthermore, and in sharp contrast to the discourse of data-driven curriculum policy-making, NAT emphasises the importance of questions as drivers of such processes. It is the question, rather than the answer, that invites participants to continued problematisation and reflection and therefore should be the driver of curriculum work at all levels. This must not be mistaken for a relativistic position denying the possibility of any kind of affirmation; rather, it is to say that any kind of affirmation has to be accompanied by an element of reflexivity, also questioning the normative assumption underpinning such an affirmation. In proposing such a reflexive approach to curriculum policy-making, I would argue that NAT offers a rich starting point for a widened, and most-needed, discussion on what is valuable to education, transcending what can be expressed as numerical data in ranking lists and league tables.