Abstract
This article explores how the concept of Universal Design for Learning constructs problems relating to education and how it constructs solutions to these problems. This is done through a five-step critical discourse analysis of the founders’ extensive website. Our analysis shows that the problems of concern relate to a discourse of barriers; ‘barriers to learning that millions of people experience every day’. These barriers are not explicitly defined but further exploration reveals that primarily traditional teaching is at fault, being too rigid and not considering students’ differences. In terms of solutions, the website offers a multitude of recommendations on how to meet students’ differences, materialised through both visual and verbal representations, in a discourse of almightiness. The vast network of actors that is mentioned on the website emphasises the far-reaching ambitions for UDL. However, we suggest that the expectations it places on teachers are unreasonable and sub-optimal for students.
Introduction
Several new pedagogical concepts have gained international interest over recent decades, including Assessment for Learning (AfL), student-centred learning (SCL), socio-emotional learning (SEL), and the Flipped Classroom (FC). We have explored the rise of such concepts in the Swedish context through earlier research which applied a ‘follow the money’ approach (Ball, 2012) to explore which concepts were popular among stakeholders and teachers in three municipalities, as evidenced by invoices on continuous professional development (CPD) accounts. Universal Design for Learning (UDL), Differentiated Teaching (DT), Clarifying Pedagogy (CP), and Pulse for Learning (PfL) emerged as popular. 1
We argue that the growing interest in such concepts is part of a tendency towards marketisation and privatisation, as they are primarily provided by commercial actors (Norlund et al., 2024). We also argue that the use of acronyms is intended to lend legitimacy to the concepts and give the impression that they are well-established.
In this paper, we focus on the concept of UDL which was launched and promoted by the American organization CAST (formerly the Centre for Applied Special Technology). CAST has its roots in architecture, working on removing physical obstacles for certain groups of people, but now presents itself as ‘a non-profit education research and development organisation that created the Universal Design for Learning framework and UDL Guidelines’ (CAST, 2024).
CAST’s policy ambitions are high; with the word ‘policy’ yielding 130 hits on the organisation’s webpage with reference to books, podcasts, and other material. The following text excerpt illustrates this: In 2006, CAST joined with several organizations to form the National UDL Task Force, an interdisciplinary coalition that advocates support for UDL in federal, state, and local policy. The Task Force has successfully advocated for the inclusion of UDL in the federal Higher Education Opportunity Act of 2008 and in various policy directives from the US Department of Education.
According to the organisation itself, the UDL concept has global reach while in Sweden the concept is recommended by two powerful policy actors; The National Agency for Education (Skolverket, 2025) and The National Agency for Special Needs Education and Schools (SPSM, 2025). Both have promoted the concept in connection with a large-scale national effort to equip teachers with generic special educational needs competencies. To date, this has engaged approximately 35.000 school teachers across Sweden. The spread and popularity of UDL is confirmed by the fact that several compulsory schools (see Lidingö stad, 2024; Pedagog Stockholm, 2025) and some universities (see Lund university, 2025; Örebro university, 2025) base their pedagogy on the concept. The popularity is also remarked upon by Murphy (2021) who refers to it as the ‘latest trend in popular pedagogy’ in his policy research note (p. 8). Any pedagogical idea which is promoted by authorities and other influential actors must, we argue, be properly scrutinised, and the aim of this study is to do this by examining the values and ideas that underpin UDL.
Previous research and previous perspectives
The UDL concept has been the focus of numerous earlier studies, most of which seem to highlight its potential. Some have highlighted its relevance to inclusion. For example, Kelly et al. (2022) concluded that applying the UDL concept can benefit student inclusion in an outdoor pedagogical setting, while Scott et al. (2019)’s intervention study showed that teacher candidates found the UDL framework helpful in making classrooms more inclusive. Levey (2023) further underscores the advantages of UDL by synthesising the results from several studies, as do Cumming and Gilanyi (2023) who report positive results on aspects such as attendance and behaviour among students with special social needs and teacher satisfaction.
Twelve peer-reviewed articles on UDL, published between 2012 and 2015, were compiled by Al-Azawei et al. (2016). Eleven of these revealed positive effects on students’ learning. However, the authors do highlight some limitations, such as that most studies were conducted in North America. It is worth noting that the authors’ conclusions seem to adhere to the conviction that ‘[t]he traditional teaching approach of “one-size-fits-all” cannot meet learner diversity in contemporary learning’ (p. 53). The authors call for further research, as does Capp (2017) following a meta-analysis of 18 studies of UDL.
We claim that there is not just a need for further research, but a particular gap in the literature relating to critical views of UDL. Capp’s (2017) analysis is one of the few exceptions. It shows that, while UDL is successful in improving students’ learning process, there is little evidence about the effectiveness of UDL in terms of learning. Capp suggests that apparently positive results regarding the latter may simply reflect methodological limitations. Moreover, Matthews et al. (2022) and Boysen (2024) share Capp´s concern about the evidential base of the concept.
Discourse analysis, method, and material
The aim of this study is to explore which values and ideas are involved when UDL advocates promote the concept and consequently construct problems and solutions related to teaching and learning.
The study is based on critical discourse analysis, a theoretical and methodological approach introduced by Norman Fairclough in which a discourse refers to a ‘way of signifying experience from a particular perspective’ (1995, p. 135), thus invoking certain values and ideas. The term ‘critical’ refers to the method’s ambition to reveal power relations.
To study the phenomenon of UDL we explored the numerous links to diverse resources housed on CAST’s main website and analysed what these reveal. The best-known resource on UDL, the UDL guidelines (UDL guidelines, 2024) were thus included. Links to online courses were excluded.
Our analysis of the website is based on a combination of Fairclough´s analytical steps (Fairclough, 1995, p. 209–210) and a modified version by Guo and Shan (2013) which has previously been applied by Levinsson and Norlund (2018), Norlund (2020), and Levinsson et al. (2022), and involves the following five steps: (1) Focus on a social problem which has a semiotic aspect. Analyse how the problem is portrayed/construed. Identify which discourse/s are involved. (2) Analyse how the suggested solution is portrayed/construed. Identify which discourse/s are involved. (3) Map the network of practices within which the problem and solution are located, and how relevant practices are potentially reorganised. Consider whether the network of practices (the social order) ‘needs’ the problem. (4) Identify potential contradictions and gaps in the material. Give space for counter-voices. (5) Reflect critically on the analysis (1-4).
In this study we will, consistent with step 1, focus on a social problem connected to education which has a semiotic aspect (as images, fonts, colours, and so on are present in the material). Semiotic resources signal something particular to the reader (Kress and van Leeuwen, 2006) and this will be extracted in the results section. For step 2, in line with Levinsson et al. (2022) we offer an analysis also of the promises made by CAST when promoting the pedagogical concept of UDL and its accompanying guidelines resource.
Results in five steps
For transparency, the results section will be presented following the five steps outlined above.
Step 1: How is the problem defined?
We find the following information about a problem high on the front page of CAST’s website: …
The problem that CAST has shouldered responsibility for is compelling: barriers are experienced by ‘millions of people’, as frequently as ‘every day’ (our italics). ‘Barriers’ are mentioned 155 times in the data, often in the phrase ‘barriers to learning’. Evidently, ‘barriers’ is the phenomenon that CAST perceives and constructs as the central problem in education today. As a metaphor, ‘barriers’ can mean many different things, but they are generally mentioned on the website without specification. Similarly, where it refers to ‘systemic’ barriers no further definition is offered. However, it is possible that systemic barriers relate to the following statement: Those barriers go by many names: racism, sexism, genderism, ethnocentrism, classism, and ableism.
It is also vague about the levels of the school system (local, national etc) at which the barriers are created, although there are some examples of where they are experienced: … the tasks of recording, organizing, analysing, and interpreting data create barriers that impede science learning for many students.
This quote highlights barriers arising from the traditional tasks performed in a specific subject area.
By following links found a couple of clicks from the home page, the reader reaches three lengthy statements that reveal the construction of problems related to students’ individual differences. We quote these in full due to their importance in highlighting the crucial values of UDL: Affect represents a crucial element to learning, and
According to these quotes, ‘learners’ can be divided into numerous subgroups based on individual differences: those who are highly engaged by novelty, those who grasp information more rapidly through auditory means, those who do not express themselves well in speech and so forth. In particular, the lower statements in bold in each verbose quote provide clues to the problem. They inform the reader that mainstream educational thinking is based on the supposedly erroneous belief that there is one optimal way for all students regardless of their differences. This assumed belief is constructed as the most fundamental barrier. Both the use of bold type and the repetition of three similar clauses signal the untouchability of the stated problem. Thus, an important feature of the problem seems to be that insufficient attention is paid to individual differences. This is explicitly re-stated in the guidelines: However, these individual differences are usually treated as sources of annoying error variance as distractions from the more important ‘main effects’.
The overarching discourse used is barriers. These are not consistently defined, examples ranging from racism to the tasks involved in studying specific subjects. However, a central claim is that traditional and mainstream teaching is too rigid and neglects learners’ individual differences.
Step 2: How are the solutions defined?
The red circle contains a plus-sign, a semiotic resource that signals addition. Clicking on this icon, the promise of an addition is logically fulfilled and the table transforms into a substantially bigger table, no longer empty of text.
Since the guidelines constitute a crucial tool within the UDL concept, it is appropriate to take a closer look at this resource. The table contains a set of measures split into three columns entitled ‘engagement’, ‘representation’, and ‘action and expression’, respectively. The green column invites the reader to ‘[p]rovide multiple means of Engagement’, the purple to ‘[p]rovide multiple means of Representation’, and the blue to ‘[p]rovide multiple means of Action & Expression’. Each column offers dot lists of urgent matters for educators to consider. Some examples are ‘Nurture joy and play’, ‘Illustrate through multiple media’, and ‘Set meaningful goals’. At the top of each column are depictions of the brain. The brain images show areas, intended to illustrate which part of the brain is activated, respectively, by efforts aiming at ‘engagement’, ‘representation’, or ‘action and expression’ (see Figure 1). The semiotic resource showing which part of the human brain is activated by each aspect of designing learning.
At the bottom of each column is a fourth category, ‘Goals’, aimed at fostering ‘Expert learners’. Each column is categorised into themes, for instance ‘Self-regulation’, ‘Language & Symbols’, and ‘Physical Action’. The vertical columns intersect with a tripartite division presented to the left labelled ‘Access’, ‘Build’, and ‘Internalize’. In total, the table provides 31 solutions as numbered ‘Checkpoints’. A click on each checkpoint leads to further recommendations. The reader thus gets specific advice relating to 142 points, 172 if subpoints are taken into account. Some checkpoints include many further points. To illustrate the checkpoints and recommendations we offer the following example from the purple column: Present key concepts in one form of symbolic representation (e.g. an expository text or a math equation) with an alternative form (e.g. an illustration, dance/movement, diagram, table, model, video, comic strip, storyboard, photograph, animation, physical or virtual manipulative).
Notably, each solution draws on multiple ideas and options.
Having presented the solutions offered by CAST to the problem of rigidness we now turn to the promises they pledge.
One click from the start page the reader can learn the objective of the organisation:
The phrase ‘a global community’ shows that the promise is far-reaching. This ambition is emphasised repeatedly throughout the webpage, through phrases such as ‘at all levels’, ‘across multiple domains’ ‘throughout the world’, and ‘by educators across the globe’.
The textual message is combined with a visual imagery. At the top of the homepage a narrow banner stretches over the page. It is dark purple with elements of blue, yellow, light green, and orange, and includes a middle line: over this line are shapes suggestive of the sun and mountains, below which there are shapes redolent of waves. The banner functions as a plate for the slogan Until learning has no limits. In combination, this suggests associations with the universe and to eternity. In this way, the banner accords with the ‘U’ in the concept of UDL, that is, Universal. Included is also imagery with round or oval, partly shadowed, shapes, reminiscent of the planets. Thus, the text resources interact with other semiotic resources to strengthen the value of the far-reaching promises. The visual and textual resources represent a promise of reach beyond the globe.
Promises are not only strengthened by references to the universe but also to a scientific anchor. The statement of CAST’s general ambition states that: We do this by helping educators and organizations apply insights from the learning sciences and leading-edge practices to educational design and implementation.
The phrase ‘from the learning sciences and leading-edge practices’ emphasises that the UDL concept is scientifically based. This adds further to the promises given, as does the fact that each of the 31 checkpoints links to reference lists, again announced with a red circle, although this time containing an arrow. For example, checkpoint 7.1 is backed up by 37 publications sorted into categories of ‘Experimental & Quantitative Evidence’ and ‘Scholarly Reviews & Expert Opinions’. The extensive lists of references enhance the impression of weightiness and legitimacy.
In sum, our analysis reveals a pedagogical concept surrounded by a range of impressive phenomena: the number of suggested pedagogical measures in UDL is extraordinary; all learners are ensured access to favourable learning conditions; the implications of the concept are of interest to an extensive group of people; the whole world (and perhaps even more than that) is involved; and the network is extensive, which makes the concept appear trustworthy and reliable. Thus, the problem discourse of barriers is, according to our analysis of solutions and promises, met and encompassed by a discourse of almightiness. UDL simply helps everyone, everywhere, at every level, and with everything.
Step 3: Network of practices and their reorganisation
The middle step in our analysis involves three stages: (a) mapping the network of actors and practices involved in the problems and solutions; (b) illustrating how relevant practices are potentially reorganised; and (c) reasoning whether the network of actors and practices ‘needs’ the problem/s.
Starting with stage (a), we note that the construction of problems and measures binds together a variety of actors and practices in networks. Under the heading Funders & Partners the website states that: CAST’s work is made possible by the generous financial support of foundations, corporations, government agencies, and individuals.
An extensive list of ‘past and present supporters’ includes the Bill & Melinda Gates Foundation, Eastern Bank, and Oak Foundation, among others. Also mentioned are school partnerships (K-12 classrooms), research organizations (e.g. SRI International), universities (e.g. the University of Kansas, the University of Michigan), government agencies (e.g. the Department of Education, the US Department Labor, public school systems), and corporations (e.g. Pearson, Google, Houghton Mifflin IBM). Having government agencies in general and the Department of Education in particular as partners could be important to CAST’s aforementioned policy ambitions. In sum, CAST not only has a varied and powerful network but is also eager to highlight it. This is consistent with the promises already noted.
It is possible to understand the network as a result of a neoliberally oriented contemporary situation in education. Decades ago, ideas moved from academia down to the local classroom in a recontextualization process (Bernstein, 2000) with only little involvement from glamourous fundraisers or commercial actors. Now, neoliberal tendencies including decentralisation and marketisation have paved the way not only for corporations but also for entangled networks in which the responsibility, ideological interest and influence of each actor is hard to identify. The possibility of influencing policy makers is simply given much latitude.
Soto and Pérez-Milans (2018) applied a similar discursive approach to studying a language implementation process in Hong Kong and showed that the policymakers behind the reform relied on arguments about liberating students from traditional teaching in favour of more flexible ways of learning. Also these authors place the UDL phenomenon within a neoliberal frame in which educational programmes have become a commodity.
Moving to stage (b), a reorganisation is taking place where practice changes from traditional approaches, based on teaching and learning as collective activities, to new approaches in which teaching is individualised and teachers cultivate as many relationships as there are students in a class (Norlund and Levinsson, 2023).
Finally, reaching stage (c), we would argue that CAST needs the problem because of its potential financial benefits. According to its 2022 Financial Report, CAST’s total revenue was $14,105,664 and its net assets at the end of the period were worth $5,789,741 (CAST, 2023). It is likely that other actors within the network also need the problem. For instance, we can suppose that business actors such as Google find it easier to sell software if students are considered to have markedly diverse needs in terms of engagement, representation, and action and expression. This supposition is supported by the fact that the UDL spokeswoman Levey (2023) highlights the need for technological support not just in the shape of devices such as computers and tablets, but also in named programmes and apps such as Kidspiration, Smart Notebook, Scene & Heard, and Smartpen. Evidently, non-traditional approaches and an emphasis on individual learning are prerequisites for commercial gain. The connection between business interests and pedagogical concepts that address learners’ individual differences and needs has been reported elsewhere (see Tomlinson, 2012).
Step 4: Contradictions and gaps
In this section, we identify potential contradictions and gaps in the CAST material and give space to counter-voices, highlighting the various critiques that can be directed towards the content and discourses that emerged from steps 1 and 2 of our analysis.
One contradiction is the organisation’s claim to scientific neutrality on the one hand and, on the other, its promotional character. CAST’s scientific credibility is bolstered by the lengthy lists of peer-reviewed references, suggesting it has a solid research base. Meanwhile, its promotional character emerges through the numerous commodities that it offers, the fact that despite being a non-profit organisation it seeks ‘a Marketing & Sales Intern’, and the grandiose language it uses.
A closer look at the reference lists reveals that a high proportion of the cited articles have indeed been published in peer-reviewed journals. A potential weakness, however, is that some of them are dated (which the text recognises) and narrower in scope than the recommendations that CAST makes. Murphy (2021) even mentions anecdotal evidence and self-publishing. Another weakness relates to claims in the purple column that ‘customizing the display of information’ is a general need, while the references linked to this point do not support this. Rather, many of the studies cited relate to children with impaired vision, for whom it should be self-evident that different ways of displaying information are needed. Also Matthews et al. (2022) note that the cited literature suffers from irrelevance in regards of the pedagogical suggestions made.
The CAST staff is introduced as including ‘experts in human development, neuropsychology, instructional design, literacy instruction, learning analytics, technology design, and more’. This seems intended to suggest that the expert base is broad and potent. The reference lists broadly correspond to the fields mentioned, though they are notably weighted towards psychology. Given that the barriers discussed relate to sociological issues such as ‘racism, sexism, genderism, ethnocentrism, classism, and ableism’, they display a significant lack of sociological research. In other words, the problems and solutions do not align.
It is logical that neuroscience is involved in the references backing up many of the checkpoints. This discipline has gained massive support in recent years, not least from policymakers as well as commercial actors. The brain simply seems to function as ‘a useful metaphor and promotional tool’ according to Boysen (2024, p. 1232). However, the relevance of combining neuroscience and education into neuroeducation has been contested both by scholars in neuroscience (Clement and Lovat, 2012; Willingham, 2009) and in education. Levinsson and Norlund (2018) from the educational discipline have noted not only a problem of commercialisation but also a risk for trivialisation of what teaching and learning is.
In teaching practice based on a UDL approach teachers are expected to not only motivate each student in different ways but also to supply each student with different material and arrange for different ways for them to display their knowledge. Meeting the expectation that teachers can provide individual solutions to each of their students across 172 aspects (i.e. the checkpoints, often specified in wide-ranging enumerations) would seem near impossible. Moreover, sustaining multiple parallel tracks during each lesson seems likely to turn teachers into busy logisticians rather than educators. Nonetheless, it is logical that every new concept must be positioned against traditional teaching – otherwise what would its contribution (to the market) be?
Meanwhile, studies in the learning sciences and sociology of education (Norlund, 2009, 2018) indicate that over-individualistic approaches like those suggested in the UDL guidelines are not benevolent from a student perspective. Rather, they may deprive students of the opportunity to learn together with their peers and take advantage of grappling with the same concepts and sharing the same questions, digging deep into different reasonings under the firm guidance of a teacher. It is troubling that CAST neglects these limitations.
A counter-voice might point out that the salient semiotic features of brains in gaudy colours found in the UDL guidelines are deceptive. In reality, brain activity cannot be isolated in this way (Gaussel and Reverdy, 2013). UDL shares similarities with the heavily criticised neuromyth of learning styles (Howard-Jones, 2014; Murphy, 2021; Norlund et al., 2024). Levey (2023), whose study we have already reported, maintains that students learn either through listening or from visual material, while the concept of learning styles erroneously maintains that people can be divided into seven, or sometimes 20 or 24 learning styles, and UDL maintains that ‘[t]he way people learn is as unique as their fingerprints’ (CAST, 2024). Thus, UDL adds to the myth by referencing immense numbers.
In contrast to our previous studies, in which we observed that writers of text went to great lengths to describe problems in drastic terms (Levinsson and Norlund, 2018; Levinsson et al., 2022; Norlund, 2020), the current study reveals a surprising absence of informative description of problems. The problems that CAST and UDL want to address become more apparent through the solutions mentioned than through discussion of problems. CAST and UDL frequently present solutions to problems that have not been previously introduced (e.g. lack of equity, flexibility, inclusion, and access). This creates the impression that CAST is less interested in going into depth about current problems or their causes and more interested in presenting solutions and promises. This matter is also reflected in the use of visual resources. Not much more than bold style is used to accentuate the problems whereas the solutions are emphasised by appealing images, promising shapes, gaudy colours, lengthy dot lists and science-resembling illustrations. A counter-voice might argue that this unbalanced focus on solutions is a promotional strategy.
Step 5: Critical reflections on the analysis
The final step includes critical reflections on the analysis conducted. One concern is periodic updates to the UDL guidelines. Our analysis was conducted using the 2.2 version which was later updated to a 3.0 version with changes in both phrasing and colour choice. However, these changes are slight and do not fundamentally alter the analysis offered in this study.
Another concern is the difficulty of doing justice to the full range of information on the website material, given its expanse. However, we have been careful to select the most essential and representative elements. We have not taken part in the online courses offered, and suggest that the content of these courses could be the subject of further research.
Concluding remarks
In this article we have studied how the influential concept of UDL portrays problems relating to education and how such problems can be addressed. Through a critical discourse analysis we identified a general problem discourse of barriers alongside a solution discourse of almightiness.
According to Fairclough (2003), the point of critical discourse analysis is its potential to reveal assumptions made by the actors involved and, by extension, how power is exerted in a particular practice. Our critical position aligns with Singh (2017)’s point, referencing the work of educational sociologist Basil Bernstein, that (new) pedagogic discourses should always be questioned: … in terms of whose interests are being served, what types of social/pedagogic relations are being constituted and what social/pedagogic identities are being formed, with what potential consequences (p. 148).
In line with Murphy (2021), we maintain that ‘[i]nstitutions of any educational level should proceed with caution before devoting significant resources to implementation of UDL’ (p. 7). In this study, we have revealed whose interests are served by UDL that UDL risks teachers facing unreasonable expectations, and students’ best interests not being served. As such, our analysis adds a critical view to the current body of knowledge which otherwise generally frames UDL positively (Capp, 2017).
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by The Swedish Research Council under Grant number 2019-03828.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.