In my recent book (Routledge 2026) I describe the WPR approach as a “new thinking paradigm”. For those new to these posts, WPR is the acronym for What’s the Problem Represented to be?, initially developed as a mode of critical policy analysis (Bacchi 2009). The argument in the new book is that WPR has important uses beyond policy analysis because, in effect, it offers a way of thinking differently, captured in the term “WPR thinking”. 

“Differently from what?” you may ask. Put simply, it mounts a challenge to the problem-solving paradigm that historically and currently dominates the intellectual and policy landscape. While there have been many critics of problem-solving, WPR offers an alternative. It challenges the presumption that problems simply exist waiting to be solved and argues instead that proposals for change (“proposed solutions”) constitute “problems” as particular sorts of problem. 

This argument should be familiar to those on the WPR list ( Welcome to the WPR Network! | Karlstad University (kau.se) and to those who have used the approach in their research. The new book shows how this analytic strategy provides openings for critical thinking in areas uncharted in earlier writing. In effect then it expands the reach of WPR thinking. New target areas include: items (such as maps or buildings), theoretical assumptions (to produce innovative literature and scoping reviews), images, media reports, and many others. 

The new book also makes a claim that a range of key terms in social and political analysis – e.g. “crises”, “issues”, “difficulties”, “matters of concern”, and others – operate as placeholders in much the same way that “problems” do. Hence, as with “problems”, they need to be displaced. The new book pursues a critique of all such reactive modes of explanation and analysis. Today’s entry offers an example to illustrate how WPR operates as a useful analytic strategy in diverse sites. 

Using WPR thinking

I was prompted to produce this entry by the following article: 

Costello, E., Ferreira, G., Hrastinski, S., McDonald, J.K., Tlili, A., Veletsianos, G., Marin, V.I., Huijser, H. & Altena, S. (2025). Artificial Intelligence in Education Research and Scholarship: Seven Framings. Journal of University Teaching and Learning Practice, 22(3). https://doi.org/10.53761/xs5e3834 

The article does not mention “What’s the Problem Represented to be?” in the text. And it does not apply the WPR questions (see Table at the end of Chapter 1 in Bacchi 2026). Yet, I have added it to the growing list of WPR applications that explicitly apply WPR ( Welcome to the WPR Network! | Karlstad University (kau.se)

Why have I done this? The authors use this quote as an epigraph: 

“What one proposes to do about something reveals what one thinks is problematic (needs to change)” (Bacchi, 2012, p. 21). 

The quote is taken from a chapter I wrote in 2012 introducing a WPR mode of analysis. In my new book (Bacchi 2026) I describe this argument as the key premise in WPR. It makes the point that WPR challenges the presumption that problems simply exist waiting to be solved and argues instead that proposals for change (“proposed solutions”) constitute “problems” as particular sorts of problem (see opening paragraph above). By placing this quote as an epigraph, the authors signal that they intend to deploy “WPR thinking” (though they don’t name it as such). What shape does this take? How does WPR thinking prove useful in their account? 

The Costello et al. article offers seven proposals for how to approach the topic of AI in Education (AIED). The authors refer to these proposals as “framings”.  In summary, these framings include: methodological pluralism; metaphors; ethnographic studies; imagining futures through fiction; humanistic groundings of AI design and development; third space professionals in research; and open education. The goal in adopting these “framings” is to open up for consideration a variety of ways of problematising AI in Education. 

Costello et al. (2025) describe their project as developing a “carrier bag” of “problem areas, approaches and framings” (p.1). The declared objective is to resist “‘hero narratives’ of technologies as weapons of domination”. Instead, the authors defend the usefulness of “small bags used in the practice of foraging”. There is a clear link here to a poststructuralist discomfort with grand narratives and close attention to site-specific conditions. 

Applying the thinking in the quote from my 2012 chapter set out in the epigraph, the seven proposals (see above) offer different ways to problematise AI. In the Conclusion, the authors explain that their intent is to produce research that “widens the conversation on AI so we see it through different lenses and frames”. In particular, the objective is to offer modes of analysis that assist researchers “to connect research on AI to the overarching aims of education itself”. We are told that disagreements among the seven contributing authors are to be expected. Indeed, “dissensus and diversity” are considered to be “useful”. 

This form of engagement resonates with WPR thinking and its commitment to continuous problematisation. However, importantly, the authors do not subject the seven proposed “solutions” (their seven identified “framings” or problematisations) to the WPR questions, as would be expected in a WPR analysis. I proceed to comment on the seven proposals, how they are elaborated and where such questioning would have been useful. To begin, I wish to emphasise the novelty of the methodology adopted in the article. 

Collective and collaborative research

The Costello et al. (2025) article offers an innovative collaborative approach to the selected topic (how AI is problematised). Each of the seven “framings” is produced by a different author, generating research that is “necessarily idiosyncratic, situated in our research and teaching practices”. The article describes their “method” in some detail:

writing was conducted in a sprint over the summer of 2025 using a shared Google doc. The first author led the ideation of an initial list of topics. The group discussed topics, agreed the final set and worked on their sections in the shared document which helped authors avoid overlap and attempt to find continuities. 

Disagreements, we are told, are to be expected (see above) because authors with contrasting backgrounds were involved in the exercise.

The question of how to incorporate collective research in WPR analyses is taken up in a recent publication by Hickman and Muir (2025). How to produce WPR as a “group exercise” is pursued in a subsequent Research Hub entry. 

AI and modes of analysis

Remembering that dissensus is to be expected, what do the seven recommended “framings” achieve in terms of innovative problematisations of AI? And where could “self”-problematisation add to the analysis? 

Methodological Pluralism

This contribution challenges the tendency to describe AI in extremes, as either “a technological utopia” or “a dystopian future of human irrelevance”. In the place of this representation, it offers to illuminate “the plurality and messiness of the ways in which technologies are used in education”. To this end it endorses methodological pluralism, including interpretive approaches, “ranging from ethnography to phenomenology to discourse analysis”. At the same time, it argues that “computational and data science methods can help researchers make sense of vast datasets of user-AI interactions”. 

As already mentioned, the seven frames in the Costello et al. (2025) article are not targeted for a WPR analysis. If such analysis had been included, there would have been room to question the assumptions underpinning methodological pluralism. This form of analysis appears in my new book in its endorsement of “paradigm talk” (Bacchi 2026, Chapters 15, 16, 17 and 18). 

Problematising through Metaphors

This contribution offers a two-pronged form of analysis. First, it suggests the usefulness of critically analysing the metaphors used to describe AI; second, it recommends using AI metaphors creatively. 

The former (examining existing metaphors), it is argued, reveals “how AI is conceptualised and how education is imagined – surfacing assumptions about pedagogy, the learner and the role of teachers”. The contribution offers the example of metaphors that anthropomorphize AI, making it more “personable” and thus “easier to be unsuspectingly peddled”. 

On the latter (promoting creative use of metaphors), the author of this section of the article puts the case that “problems already given may have to be reframed”. Here the author directs attention to the vexed question of plagiarism, “perhaps an issue soon to become a non-problem due to GenAI”. In a clear example of WPR thinking, the author describes how “the ‘solution’ presents itself in the shape of a burgeoning market of AI detection platforms”. This problematisation is described as producing “a ‘Bootleg” industry of fabricated solutions to fabricated problems”. The usefulness of WPR thinking in relation to the topic of GenAI policies in higher education assessment is demonstrated in several recent articles (see Groves and Nagy 2025; Luo 2024; Mochizuki, Bruillard & Bryan 2025). 

Ethnographic Study of AI in Education

This contribution suggests the benefits of producing an ethnography of “technology hype itself”. It asks why “we can look back on previous technology hypes so critically, while at the same time convincing ourselves that this time it is going to be different”. What is needed, in this account, is “deep, extended observation and participation that tells us something about our own tendencies to adopt technologies so uncritically” and “collective autoethnographies” that focus on “localised practices”. 

As with the section on methodological pluralism, the use of ethnography to problematise AI is not subjected to a WPR analysis. My new book (Bacchi 2026) contains a chapter (Chapter 18) that undertakes this task. 

Imaging Futures through Fiction

Education fiction is promoted as a way to prompt reflection on the uses of AI in education, and to allow us “to act in the present by imaging possible futures”. Fiction is described as a “powerful practical strategy for both teaching and learning that can foster critical and technoskeptical thinking in students”. Education fiction becomes a creative tool “with which to discuss complex and often difficult topics around AI’s influence”. 

In the recent IPPA conference in Chiang Mai (2-4 July 2025), Laura Bea, in a special WPR seminar, presented a paper titled, “Can fiction help us to rethink public policy on violence” (Critical Policy Conversations, https://criticalpolicy.co.uk/icpp7/). Related issues are raised in a special Research Hub entry (29 Nov 2022) on “Sociotechnical imaginaries and WPR: Exploring connections” (see also Rahm and Rahm-Skågeby 2023). The task becomes examining the problematisations in this material to see what it can produce as useful political analysis.

Humanistic Grounding in AI Design and Development

This contribution to the Costello et al. (2025) article stresses the need for a humanistic ethics in AI research, which shifts “the focus from the outcomes that AI can achieve to the procedures through which it does”. The declared objective is to create “user-centred solutions that address the diverse needs of learners”. The question according to this author becomes: “Which humanistic principles should be considered when designing and developing AI-powered systems in education?”. 

The entire Costello et al. (2025) article reflects strong humanistic principles. The opening sentence reads: “The words you are reading right now were written by a human being”. The closing sentence to the article states emphatically that “the messiness of the collaborative activities of both teaching and research” are “never entirely determinate practices and always have passionate human beings at their heart”. 

There appears to be no place in the article to query humanist assumptions. WPR thinking would create space to consider how subjects are conceptualised in this account. 

Third space professionals in AI educational research

The author of this section stresses the need to broaden the focus on “teachers” in AI educational research to include “other actors in universities who contribute to learning and teaching”. These actors include “learning designers, academic developers and learning technologists”. These professionals are considered to “add considerable value to how the AI research agenda is conceptualised, designed and enacted”.

Entanglement of Open Education and AI

This contribution emphasises the need to consider the relationship between AI and Open Education as an “entangled pedagogy”. On the one hand, the author argues that the “potential of AI to support digital public goods” ought to be recognised. On the other hand, s/he highlights the challenges AI raises to Open Education in the form of “algorithmic bias, digital divide expansion and over-reliance on proprietary AI models”. 

The example of open licensing (e.g. Creative Commons) is put forward to illustrate the challenges and tensions in this “entangled” relationship. This author makes the case that the whole meaning of “open” in Open Education is put into question because generative AI can generate content “without reference to the aggregated sources used to produce the output”. The author expresses the need for more research “on this evolving notion of openness in the context of education”. In other words, s/he wishes to problematise “openness”, a topic that could be explored using WPR.

Conclusion

The structure of the Costello et al. (2025) article reflects what I call “WPR thinking” in several ways but stops short of applying the WPR questions to its own proposals (see Process 7 in WPR template; Chapter 1 in Bacchi 2026). It offers seven “lenses” through which to consider how AI is conceptualised. On occasion it also uses examples of application (e.g. on plagiarism) that closely follow the WPR mode of inquiry – starting from proposals (AI detection platforms) and working backwards to identify problem representations. 

The focus on competing entry-points to studying AIED with the expressed goal of widening the conversation on AI reflects the WPR intent to explore complexity in social relations. Given the diverse backgrounds of contributors, understandably, the uptake of WPR is uneven. “Self”-problematisation (Process 7) could have provided a means to surface these contrasting perspectives and their effects, an argument pursued in the next entry. 

References

Bacchi, C. 2009. Analysing Policy: What’s the Problem Represented to be?Frenchs Forest, NSW: Pearson Education.

Bacchi, C. 2012. Introducing the ‘What’s the Problem Represented to be?’ approach. Engaging with Carol Bacchi: Strategic interventions and exchanges, 21-24. in Eds Blestsas, A., & Beasley, C. Engaging with Carol Bacchi. Strategic interventions and exchanges. University of Adelaide Press. https://doi.org/10.1017/UPO9780987171856.003 

Bacchi, C. 2026. What’s the Problem Represented to be? A new thinking paradigm. New York: Routledge. 

Costello, E., Ferreira, G., Hrastinski, S., McDonald, J.K., Tlili, A., Veletsianos, G., Marin, V.I., Huijser, H. & Altena, S. (2025). Artificial Intelligence in Education Research and Scholarship: Seven Framings. Journal of University Teaching and Learning Practice, 22(3). https://doi.org/10.53761/xs5e3834 

Groves, A. and Nagy, V. 2025. Crime or Failure of Integrity: What is the Problem of Contract Cheating Represented to be in Australia? Higher Education Policy,https://doi.org/10.1057/s41307-025-00402-6

Hickman, M. E. & Muir, R. (23 Sep 2025): Integrating co-analysis and researcher reflexivity into Bacchi’s ‘what is the problem represented to be?’ framework: A cervical screening case study, Critical Policy Studies, DOI: 10.1080/19460171.2025.2561142 

Luo, J 2024, “A critical review of GenAI policies in higher education assessment: a call to reconsider the ‘originality of students’ work “, Assessment & Evaluation in Higher Education, DOI: 10.1080/02602938.2024.2309963 

Mochizuki, Y., Bruillard, E. & Bryan, A. (21 May 2025): The ethics of AI or techno-solutionism? UNESCO’s policy guidance on AI in education, British Journal of Sociology of Education, DOI: 10.1080/01425692.2025.2502808 

Rahm, L. and Rahm-Skågeby, J. 2023. Imaginaries and problematisations: A heuristic lens in the age of artificial intelligence in education.  British Journal of Educational Technology, pp. 1-13. DOI: 10.1111/bjet.13319