In this fourteenth instalment of the ‘Let’s Talk Open Science’ series, the CCSD proposes putting the concept of ‘openness’ to the test through the lens of ‘openwashing’, or the subversion of openness in scientific publishing. How can a widely accepted and valued concept mask practices that remain partly closed, proprietary or commercial?
This webinar featured Chérifa Boukacem-Zeghmouri, a professor of information and communication sciences at the Elico laboratory (Claude Bernard Lyon 1 University), and Hans Dillaerts, a senior lecturer in information and communication sciences at the Lerass laboratory (Paul Valéry Montpellier 3 University). Their work is part of the ROR² project (Open and Responsible Research x Research on Research), which is funded by the ANR as part of the RESO programme (Research on the Practices and Challenges of Open Science).
A disconnect between rhetoric and practice
The webinar aimed to examine the conditions under which open access is possible, rather than to condemn it. Questions were asked such as: who defines ‘openness’, which mechanisms make it effective, and which, on the contrary, undermine its meaning?
Guest researchers proposed considering ‘openwashing’ as a phenomenon of decoupling between the professed values of open access (such as democratisation, accessibility and the circulation of knowledge) and the resources actually deployed, such as publication charges (APCs), commercial exploitation, proprietary platforms and opportunistic use of the ‘open’ label. This notion of decoupling refers to situations in which officially stated discourses and standards do not fully correspond to implemented practices, without necessarily implying a systematic intention to deceive.
The term ‘openwashing’, which first emerged in the late 2000s within the contexts of free software and open data, did not come to be used in the realm of scientific publishing until the early 2020s. Although criticism of the commodification of openness (APCs, strategies employed by major publishers and predatory journals) had already existed prior to this, it had not yet been formalised under this label.
The speakers emphasise that their definition is not normative: their aim is to objectively analyse the mechanisms at play, not to judge the practices of those involved. Openwashing is described as a form of low-cost deviance which does not manifest as spectacular fraud, but rather becomes gradually embedded in professional routines to the point where it sometimes becomes invisible.
This research is the subject of a preprint co-authored by the two speakers and Guillaume Silhol, which is available on HAL as ‘Legitimations and Subversions of the Open: Towards an Analysis of Openwashing in Scientific Publishing’.
Practical examples: from editorial rebranding to package-based flipping
How, then, can ‘openwashing’, which is still being analysed as part of the project, manifest itself?
One example is academic social networks that are transforming into publishers, as illustrated by Academia. After considering the idea of reselling articles submitted by researchers but not following through on it, the platform launched a catalogue of gold open-access journals, which now comprises around twenty titles. While these journals share free access to articles with open access, they do not incorporate the other aspects of open science, as their editorial boards often originate from the platform itself and draw on researchers on the fringes of the mainstream publishing system.
The rebranding of existing journals by adding the word ‘open’ to their titles, accompanied by a broadening of their editorial scope. This phenomenon enables certain publishers to target new markets, particularly in Asian or Eastern European countries, where APC rates are sometimes lower than those in Western European markets, and to attract some of the articles rejected by more prestigious journals.
The repurposing of ‘flipping’. Originally, ‘flipping’ referred to a journal’s transition from a subscription model to an open access model supported by its scientific community. The speakers demonstrate how this movement, born out of an effort to reclaim editorial control, has been co-opted by certain major publishers, who are now switching entire portfolios of journals to gold open access as soon as a certain percentage threshold of open-access articles is reached, whilst imposing targets on editorial teams to increase the number of articles published.
Finally, there has been a proliferation of non-personalised solicitations sent by otherwise legitimate publishers (invitations to submit an article, to join an editorial board, to become an editor-in-chief, etc.), whose rhetoric bears a disturbing resemblance to that of predatory journals. In this regard, a comparative study drawing on the historical analysis of the ‘confiscation of language’ has been proposed, to examine how the vocabulary of openness can be stripped of its original meaning for commercial gain.
What the researchers themselves say
The initial findings from the focus groups conducted as part of the ROR² project (nine of the originally planned ten have been carried out to date) show that researchers draw on several different conceptions of openness, sometimes without being fully aware of it. These conceptions are rooted in the values of science, driven by public and institutional policies, or shaped by platforms. This diversity is accompanied by contradictory demands, such as publishing in open-access journals while also targeting those with a high impact factor. This forces researchers to adjust their strategies on a case-by-case basis depending on the pressures they face at the time.
Private platforms are the most successful in reconciling these different perspectives within their commercial discourse, to the extent that they have become an implicit benchmark in the way researchers talk about their open access practices.
Open questions for the next stage of the project
The discussion at the end of the session provided an opportunity to explore several related issues. These included the lack of reliable forecasts regarding large-scale flipping due to a lack of research on the subject, the precarious situation of early-career researchers (identified as one of the main obstacles to confidently adopting open practices) and the role of generative artificial intelligence. The latter is not viewed as an aggravating factor in itself, but rather as a technology whose contribution to the phenomenon — whether positive or negative — depends on how it aligns with the transparency and reproducibility principles inherent in open science.
Work on the ROR² project is ongoing, with a particular focus on analysing a corpus of journals rebranded as ‘open’ and using data from the Harbinger project.