The platform had not arrived, but the school had already moved

A school can reorganise itself around a technology before the technology is fully in use. Meetings multiply. Existing subscriptions are reconsidered. Training begins. Staff start to describe themselves as preparing for an unavoidable future. By the time a platform becomes ordinary, many of the choices that made it ordinary have already been made.

This is the most important insight in Toon Tierens, Mathias Decuypere, Sigrid Hartong and Samira Alirezabeigi’s open-access study of digital infrastructure in four Flemish schools. First published online in October 2025 and now included in the September 2026 issue of the European Educational Research Journal, the paper asks how infrastructure matters affectively at the level of a school—not only when an individual uses an interface, but when promises, pressures and reputations begin to organise what a whole institution feels able or obliged to do.

An atmosphere is not just a mood

The authors use “atmosphere” to name a collective orientation that is felt across an organisation without being reducible to any one person’s emotion. Their argument is that digital infrastructures generate such orientations because they arrive as more than tools. They bring standards, subscriptions, devices, training, data arrangements, imagined futures and relationships with large providers.

The study draws on three years of ethnographic work, from 2021 to 2024, across one primary and three secondary schools in Flanders. The first author attended meetings, made regular visits and spent more intensive periods in each school. The empirical material consists mainly of field observations and statements from teachers, school leaders, ICT coordinators, care coordinators and administrative staff. Planned conversations were recorded; spontaneous conversations were documented in fieldnotes.

Two atmospheric formations organise the analysis. The first is anticipation. In two schools, ChatGPT produced anxiety and excitement before its organisational role was settled. Staff changed assessment practices, formed AI groups and prepared for integrations that had not yet materialised. One school reduced subscriptions to other learning platforms because the expected combination of ChatGPT and Moodle appeared to change what would soon be possible.

The second is legitimation. In the other schools, Google’s ecosystem and Smartschool’s integration with Microsoft did not merely promise functionality. They made a digital direction feel credible, modern and secure. The schools “nested” more deeply through devices, certifications, workshops and professional routines. The paper’s striking claim is that lock-in can be desired: dependence may feel like assurance when a familiar ecosystem seems to validate the school’s identity and future.

What the ethnography makes visible

The paper succeeds because it follows organisational movement that a conventional technology-adoption study could miss. A survey administered after implementation might record positive attitudes, concern about data or frequency of use. It would struggle to show how a vendor presentation, an API announcement or the promise of interoperability changes conversations before there is a stable object to evaluate.

The cases also resist a simple account of coercion. Staff are not portrayed as passive victims of technology companies. They worry, hope, disagree and actively invest. A teacher’s concern about data concentration is heard, but colleagues treat dissent as impractical because affiliation with a recognised ecosystem has come to signify being a good school. That episode shows how legitimacy can narrow debate without anyone formally prohibiting disagreement.

The duration of the fieldwork is another strength. The study can connect apparently small moments—an impromptu conversation, a workshop, a changed subscription—to longer organisational trajectories. Its ethics approval, consent arrangements, funding and analytical purpose are stated clearly. The authors are also careful not to claim statistical representativeness.

Where atmosphere risks explaining too much

The concept’s reach is also the paper’s main vulnerability. The conclusion proposes that infrastructures always matter atmospherically for schools. If every organisational response to infrastructure counts as atmospheric, it becomes difficult to identify what evidence would show that atmosphere is not doing explanatory work in a particular case.

The empirical analysis selects moments of intensity because they make atmosphere visible. That is defensible for an interpretive study, but it creates a selection problem. Quiet continuity, indifference, unresolved disagreement and abandoned initiatives may be less noticeable even though they could challenge the impression of a school-wide orientation. The paper acknowledges variation, yet it offers limited detail about how observations were coded, how contrary episodes were sought, or how the four researchers tested alternative interpretations of the fieldnotes.

There is also a difference between documenting shared language and demonstrating a shared atmosphere. Leaders and enthusiastic teachers may be unusually audible in meetings. Staff who comply for practical reasons, avoid discussion or work around a platform could experience the organisation differently. Because spontaneous conversations were written rather than recorded, the analysis necessarily depends on the fieldworker’s judgement about which expressions, gestures and tones were significant.

A more ordinary explanation may account for some findings. Schools facing uncertainty often imitate organisations they regard as successful, choose reputable suppliers, reduce coordination costs and seek a coherent identity. Institutional isomorphism, procurement constraints and bounded rationality could explain movement towards Google or Microsoft without requiring a distinct atmospheric mechanism. The paper is most persuasive when atmosphere adds temporal and affective detail to those explanations; it is less persuasive when it appears to replace them.

An Autopoietic Ecology reading

An Autopoietic Ecology reading begins with the difference between adopting a tool and reorganising the conditions of future action. Training, subscriptions, hardware, standards, records and professional identities can make one next step easier while making alternatives harder to sustain. The school is not simply persuaded once. Repeated investments help reproduce the arrangement through which later choices become possible and intelligible.

The article’s account of nesting connects closely with the EdgeLab Research Programme’s concern with agency, institutional authority and repair. Interoperability becomes more than a technical property when it is translated into evidence that a school is credible. A vendor’s ecosystem then mediates meaning: compatibility comes to signify safety, professional competence and readiness for the future. This is an EdgeLab interpretation, not a claim made in the paper.

AE adds a question the article only partly develops: which capacities are renewed and which are depleted? A platform may increase staff confidence, reduce fragmented logins and enable new teaching practices. At the same time, it may weaken the capacity to exit, retain local knowledge, contest data use or keep a non-proprietary alternative viable. Continuation is therefore not automatically evidence of organisational flourishing. A school can become more operationally coherent while becoming less able to revise the conditions on which that coherence depends.

The connection should not be overstated. “Atmosphere” and “ecology” are not equivalent concepts, and the study does not demonstrate autopoiesis or AE. Its evidence concerns how particular schools orient themselves towards infrastructure. The value of the comparison lies in the questions it opens about recursive dependency, not in relabelling the authors’ analysis.

What schools could examine before nesting

The practical implication is not that schools should reject integrated platforms. Fragmented systems impose real labour and can make support unreliable. The better response is to examine the atmosphere as part of the decision rather than treating enthusiasm, urgency or reassurance as neutral background.

• Record what problem the platform is expected to solve and which claims depend on a future capability rather than present evidence.
• Give dissent a defined route into procurement and review, including concerns about data, workload, pedagogy and dependence.
• Test reversibility: data export, interoperability, contract exit, retained expertise and the cost of restoring an alternative.
• Revisit the decision after implementation. Ask which promised capacities appeared, which did not, and what the school can now do less easily.
• Compare the atmosphere around the provider with evidence about actual classroom and organisational outcomes.

What could challenge this interpretation?

A stronger test would follow schools longitudinally from anticipation through procurement, routine use, failure and possible exit. It would deliberately sample quiet and dissenting staff, compare fieldnotes with interviews and documents, and examine schools facing similar policy pressures that choose different infrastructures. If the same patterns are adequately explained by price, technical compatibility and formal authority, the atmospheric account should be narrowed.

Evidence could also challenge the AE extension. Schools might nest deeply within a proprietary ecosystem while preserving genuine data portability, plural expertise, effective challenge and an affordable route out. Conversely, a nominally open system might exhaust staff time and make repair impossible. Those findings would show that provider concentration is not a sufficient measure of ecological constraint; the actual capacity to question, adapt and continue differently must be observed.

Tierens and colleagues make a subtle organisational process visible: technologies can acquire practical authority through anticipation and reassurance before their effects are settled. The next step is to specify when an atmosphere changes a decision, whose experience it organises, and what allows a school to revise the future it has already started to inhabit.

Reference: Tierens, T., Decuypere, M., Hartong, S., & Alirezabeigi, S. (2026). “Feeling infrastructures: How digital technology matters atmospherically for schools”. European Educational Research Journal, 25(5), 723–740. First published online 1 October 2025; issue published September 2026. https://doi.org/10.1177/14749041251377119.

Cambridge EdgeLab is an independent research initiative. This review represents Steven Watson’s critical interpretation and does not imply endorsement by the article’s authors, the University of Cambridge or any other institution.