Companion post to Wolfgang Kaltenbrunner’s "Research culture as a politics of uncertainty"

If research culture is a politics of uncertainty, then someone has to organise that politics. In a recent post, Wolfgang Kaltenbrunner proposed understanding research culture as a way of distributing uncertainty across research systems: who can afford to take risks, whose research problems are easier to pursue, who bears the costs. This is a powerful framing. But it raises an obvious next question, which we set out to answer empirically: how, and by whom, is that uncertainty distributed in practice, and through what systems?

Our answer, based on a comparison of Spain and Lithuania, is that a large part of the work is done by something rarely discussed in research culture debates: publishing infrastructures. These two countries show, almost as a natural experiment, that similar national policy pressure produces very different outcomes depending on whether a domestic publishing infrastructure survives to absorb this pressure.

Same pressure, two outcomes

Spain and Lithuania are both semi-peripheral European systems under the familiar pressure to internationalise: to publish in prestigious, core-based journals as the measure of success. Both saw the rapid arrival of high-volume open-access publishers such as MDPI. On paper, the incentive was identical. The outcome was not.

In Spain, a resilient ecosystem of university- and society-run journals (many in Spanish, and thus reaching readers across Latin America) remained standing. Spanish researchers could treat new open-access outlets as one more option: a portfolio addition layered onto domestic journals and traditional international ones. When MDPI’s share climbed to roughly a third of national output, the system had the resources to push back; after the national agency ANECA delisted several titles, that share fell to around 13%. Uncertainty remained spread across multiple credibility circuits.

In Lithuania, national evaluation policy had spent two decades gradually marginalising domestic journals, whose share of the country’s Web of Science output fell from a peak near 52% to about 7% by 2024. With the domestic circuit devalued, researchers had nowhere to distribute their bets. New publishers became not an addition but a systemic substitute, replacing an infrastructure that policy had hollowed out. What looks like individual strategic choice is, seen at the system level, a constrained response: uncertainty concentrated onto a narrow set of external, commercial pathways.

We call the first outcome autonomous coexistence and the second dependent displacement. The difference between them is not researcher virtue or national temperament, but whether a community-managed publishing infrastructure was allowed to keep functioning. These configurations point to a more general distinction in how systems adapt to shared pressures: whether new publishing models extend an existing repertoire of options or come to replace previously available pathways altogether.

Why infrastructure, not just norms

This is the point we seek to add to the politics-of-uncertainty framing. Uncertainty is never encountered in a pure form; it is mediated. Publishing infrastructures are one of the most concrete ways in which this mediation is organised. They define the available pathways through which work can be made visible and credible, and in doing so they decide whether uncertainty can be spread across options or gets funnelled into dependence.

That means research culture is not only about norms, working conditions, or good intentions. It is also an infrastructural condition. Evaluation systems do not merely respond to uncertainty; they systematically redistribute it, and they do so through the mundane machinery of journal lists, indexing databases, and incentives. This also reframes directionality: by privileging particular formats and databases, evaluation systems stabilise some research trajectories and quietly render others less viable.

What follows

If the above is true, then protecting and governing publishing infrastructures is not a technical footnote to the research culture agenda; it is part of the agenda. A country that dismantles its domestic scholarly publishing, as has been argued before, does not simply lose a few journals; it loses the buffer that lets its researchers distribute uncertainty on their own terms. Reforming metrics alone may not be enough if the infrastructures that give reform something to land on have already been eroded.

This connects directly to a broader conversation about metascience and democracy: about illuminating the uncertainties we tend to neglect. Bringing infrastructure into that discussion grounds debates about uncertainty, directionality, and democratic engagement in the material organisation of research systems. The distribution of uncertainty is not neutral. It is designed, and crucially, it can be designed differently.

The full study is available to read: Dagienė, E. & Aibar, E. (2026). "Publishing infrastructures in the semi-periphery: how research assessment shapes the research output of Spain and Lithuania." Research Evaluation, 35(1), rvag042. https://doi.org/10.1093/reseval/rvag042

Header picture by chi_margo on Unsplash
DOI: pending