Taking Publishing Back
Denis Bourguet and Thomas Guillemaud on Peer Community In, diamond open access, and why scholarly evaluation should be controlled by researchers
My first interview for Academic Renewal is with Denis Bourguet and Thomas Guillemaud, co-founders of Peer Community In. I am delighted to begin with them. PCI has become one of the most interesting and constructive developments in scholarly publishing: a researcher-led system for the review, recommendation and dissemination of scientific work, without the usual dependence on commercial journals and article processing charges.
I should declare an interest. I am a strong supporter of PCI, and I have known Denis and Thomas for some years. PCI puts peer review and recommendation back in the hands of researchers, and does so without asking authors or readers to pay. At a time when so much academic publishing has become expensive, opaque and driven by volume, PCI offers a different model: community-based, rigorous, transparent, public-spirited — and diamond open access. The latter means that your papers are published free of charge. Amazing, eh? In my experience, the community-led nature of PCI also brings another benefit: thoughtful, serious reviews from referees who understand the work and care about improving it.
That makes Denis and Thomas exactly the right people with whom to begin this interview series. Their work speaks directly to many of the themes of Academic Renewal: how science is assessed, who controls the means of publication, how public money is spent, and how researchers might rebuild systems that serve knowledge rather than markets.
For more on PCI click here.
PBR: What problem was PCI created to solve?
When you founded Peer Community In, what was the precise failure in scholarly publishing that you wanted to address? Was it cost, quality control, publisher power, speed, transparency, or something about who owns the process of scholarly evaluation?
PCI: At the beginning, the main motivation for creating Peer Community In was the cost of scientific publishing. Many countries simply cannot afford to publish their research or to access scientific articles. But even for well-funded institutions, the costs are huge. What we found particularly shocking was that a significant part of the money coming from public research budgets ended up as profits for a handful of large publishing companies. We thought this money would be much better spent on research itself.
This reason is still valid today, but over the years other issues have reinforced our motivation to develop PCI.
One is the quality of published science. We are increasingly aware of problems such as poor reproducibility, false positives, questionable research practices, and the publication of papers that should not have passed peer review. We believe that peer review should be organized by the scientific community, with the objective of improving science rather than serving commercial interests.
More generally, we think scientific publishing should be in the hands of the scientific community. Scientific societies have an important role to play, but many of them have become financially dependent on agreements with large commercial publishers. They receive part of the publishing revenue, which they often use to support their activities. We understand this, but from the perspective of the research community, it is not an efficient system. The money all comes from research budgets, only a small fraction goes back to the societies, and the largest share becomes profit for the publishers.
We think the scientific community should either work with publishers that charge fair prices or, even better, organize its own publishing system. Evaluating and disseminating science are core scientific activities, and researchers should keep control over them.
PBR: What does PCI show about the true cost of publishing?
One clear feature of PCI is that it separates peer review and recommendation from commercial journal publication. What does this reveal about what actually costs money in scholarly communication, and what publishers are charging for? Perhaps you could also comment on what it costs PCI to operate and the financial needs going forward to remain diamond access.
PCI: One of the interesting things PCI shows is that the actual cost of publishing scientific articles is much lower than many people imagine. By way of full transparency, which is something that we pride ourselves on, a post on PCI’s finances can be viewed here; and another on article costs is here.
Today, most of PCI’s budget is not spent on publishing articles. It is spent on developing and improving the infrastructure that manages peer review and recommendations (i.e., the back office of our websites) and on making researchers aware that PCI exists and encouraging them to use it. The second cost remains relatively small - a few thousand euros per year -and should decrease over time as PCI becomes better known. We also expect the cost of developing the website to decrease once the new version of the website we are currently building would be completed.
The operational cost of publishing itself is actually very low. Managing accepted articles, formatting them, maintaining the websites, and making the articles available online costs only a few hundred dollars per article.
What’s interesting is that the two largest costs we currently have - developing the infrastructure and promoting PCI - are costs that large commercial publishers no longer have. Their platforms are already built, and their journals are already well established. Yet they continue to charge very high prices.
So, in our view, PCI demonstrates that running a high-quality publishing system does not require the levels of expenditure that are often claimed. The real challenge is building the infrastructure and changing researchers’ publishing habits. Once that is done, the cost of operating a diamond open-access system is relatively modest.
PBR: How should scientists think about article processing charges (APCs)?
Many researchers have come to treat APCs as unavoidable, even when they reach several thousand euros. From the perspective of PCI, when is an APC justified, and when is it simply rent extraction from public money and unpaid academic labour?
PCI: We don’t think the real question is APCs versus no APCs. The real question is whether the fees reflect the actual cost of publishing.
In our view, APCs of a few hundred euros, and probably up to around one thousand euros, can be justified. The exact figure is open to discussion, but we know that high-quality publishing can be achieved at this cost. Several publishers, including commercial ones such as Copernicus and Pensoft, show that this is possible. They provide publishing services at prices much closer to the actual cost of publication, and act as genuine partners of the scientific community.
Conversely, non-profit status does not automatically make pricing acceptable. Some non-profit publishers also charge APCs that many researchers would regard as excessive.
The problem is that APCs often seem to increase with the prestige of the journal rather than with the actual cost of publication. Indeed, several studies have shown that APCs are positively correlated with journal reputation, typically measured by impact factor. This strongly suggests that authors are not paying for additional publishing services. They are paying for the prestige associated with the journal. In other words, publishers are able to capture a prestige rent that has largely been created by the scientific community itself.
That is the real problem. Researchers write the papers, review them, and often edit the journals without payment, while public research funds are used to pay APCs that far exceed the real cost of publishing.
PBR: Why have academics tolerated the current system for so long?
Scientists write the papers, review the papers, edit the papers, sit on editorial boards, and then pay publishers to make the results public. Why has the scholarly community allowed this arrangement to become normal?
PCI: There are several reasons.
First, many researchers simply do not know how the publishing system works or how much it costs. In many cases, they never see the bill. APCs are often paid by the corresponding author, covered by research grants, or, increasingly, prepaid through transformative agreements between publishers and universities. As a result, many authors publish without even knowing how much their paper costs. Even when they do know, they often think, “The money has already been paid, so I might as well use it.” This is a classic collective action problem.
Another reason is that publication costs remain small compared with the total cost of doing research. If a project costs hundreds of thousands of euros, paying a few thousand euros to publish the results may seem almost negligible.
But we think the deeper reason lies in the way scientific careers are evaluated. Since the 1980s, research systems have increasingly emphasized competition, grant funding, journal prestige, impact factors, and the “publish or perish” culture. Researchers are naturally encouraged to maximize the visibility and prestige of their publications because their careers depend on it.
The problem is that this gives publishers enormous market power. They own journals with established reputations, and researchers feel they have little choice but to publish in them. As long as scientific assessment continues to rely heavily on journal prestige, many researchers will continue to accept a system that they privately recognize as inefficient or even unfair.
PBR: What makes PCI credible?
Sceptics might ask why a PCI recommendation should be trusted in the same way as publication in a recognised journal. What are the safeguards that ensure rigour, fairness, transparency and accountability in the PCI model?
PCI: PCI is credible because it is built by researchers. The people making editorial decisions are active scientists whose primary objective is the quality of the science, not the commercial success of a journal.
One of the main strengths of PCI is transparency. The reviews, and actually the complete editorial correspondence are all public when we accept and recommend a preprint. This creates a strong incentive for recommenders to ensure that the evaluation is thorough. If a preprint were recommended after only a superficial review, everyone would be able to see it.
We also pay particular attention to conflicts of interest. Recommenders and reviewers must not have conflicts of interest with the authors or with the content of the preprint being evaluated. PCI also forbid the submissions of preprint with financial conflict of interest.
In several PCIs, such as PCI Ecology and PCI Evolutionary Biology, some evaluated preprints also undergo additional checks by data and code editors. For empirical studies, we ask that authors justify their sample sizes, and we have created a PCI specifically devoted to the evaluation of Registered Reports, which help improve the robustness of research before the results are known.
Overall, we would say that PCI is credible because it makes the peer-review process more transparent, more accountable, and more focused on scientific quality.
PBR: Can PCI scale without becoming what it was created to challenge?
Many reforms begin as community-led alternatives and then become professionalised, bureaucratised or captured. What would PCI need in order to grow while preserving its original principles?
PCI: We don’t think PCI needs to change to scale. It simply needs more submissions. In fact, most PCIs are already receiving an increasing number of submissions. Based on the current figures, we expect the total number of submissions this year to be roughly twice that of 2025.
The infrastructure is already in place, the editorial process works, and our community continues to grow. Of course, we need sustainable funding to maintain and improve the platform, but growing does not require becoming more commercial or more bureaucratic.
PCI is ready. What we need now is wider adoption by the research community.
PBR: What should funders and institutions do differently?
If funders and universities genuinely wanted to support responsible publishing, what specific changes should they make to assessment, APC policies, indexing, repository use and recognition of peer review?
PCI: First, funders and institutions should focus much more on the quality and robustness of the science they support, rather than on publication metrics.
Second, they should stop reimbursing excessive APCs. Researchers should remain free to publish where they wish, but public funds should not cover APCs above a reasonable threshold—for example, around €500 per article.
They should also encourage researchers to publish in diamond open-access journals or in journals that charge reasonable APCs, and to contribute as reviewers and editors to these community-driven initiatives – see for example the DAFNEE database.
Finally, research assessment must change. Institutions should place much greater emphasis on the quality of scientific contributions than on the number of publications or journal-based metrics such as the impact factor, whose limitations are well known. Initiatives such as DORA and CoARA are important steps in that direction and must be signed and their recommendations used by institutions.
PBR: Does journal prestige remain the central obstacle?
Even when better models exist, researchers still feel pressure to publish in recognised journals with established brands. How much of the publishing problem is really a problem of research assessment?
PCI: Research assessment is half the problem. The other half is that researchers are intrinsically, individually, sensitive to prestige, whatever its basis. It can be the journals in which they publish, the number of articles they published, the number of citations they obtained, their h-index, their Erdős number, the prizes they receive, the position they get, the congresses they are invited to, etc. All these elements are logically based on the results that are obtained by the researchers and therefore they tend to twist the results: Results are more often positive than they should be, they’re more shiny and greater than they should be, effects are larger than they should be, etc… This would be true even if we neutralize the journal prestige effect.
In a word, it’s the individual prestige measured in a multidimensional scale based on notoriety. Remove the journal prestige, and researchers will still seek something else like medals or another shiny toy. Unfortunately, this new toy will probably also bias the results toward false positives, over estimation of effects and bad science.
A solution may be to modify the notoriety scale so that it would be aligned with the quality of science. Some people already think about it, for example by proposing the TOP factors, or the PQRST scale, quality badges on publications. John Ioannidis, in a conference at UC Davis in 2019, proposed bold quantitative metrics to evaluate researchers that would be a positive function of virtuous research practices: see this YouTube video (@ 1h, 08 min, 29 sec).
PBR: What would success look like?
If PCI and related models were successful over the next decade, what would scholarly publishing look like? Would journals still exist? Would APCs disappear? Would peer review become more open, more accountable, or more community-owned?
PCI: Even if PCI became a major success, publishing several thousand articles per year, the scholarly publishing system would still be largely dominated by journals. After all, millions of scientific articles are published every year. It would take dozens, or even hundreds, of initiatives like PCI to transform the system.
What we hope is that PCI will help accelerate the development of diamond open access. Ideally, more diamond journals will be created, and existing society journals will either become diamond journals or partner with publishers that charge reasonable APCs.
As we said in our answer to a previous question, our goal is not necessarily to eliminate APCs, but to ensure that they reflect the real cost of publishing rather than the prestige of a journal.
We also hope PCI will contribute to making peer review more transparent. Publishing peer reviews, as we do when an article is recommended, encourages thorough evaluations and gives reviewers credit for an important part of their scientific work. This could also help address the growing difficulty of recruiting reviewers.
Finally, if PCI succeeds, it will demonstrate that a community-led, diamond, high-quality publishing model can work at a large scale. We think that would be an important message for the future of scholarly publishing.
We also have a secret dream: Registered Reports — a format already in place at PCI — would become widely used and even the norm. Evaluations of articles would be made by journals and other evaluation platforms before obtaining the results. Questions, hypotheses and methodology would be evaluated before making surveys, experiments and field work, and this evaluation would be the basis of the acceptance of the articles. Works would not be evaluated based on the results but on the questions (including their importance) and methodology. This would select for a better quality of science, for many reasons (better methodologies, more reproductions, less positive result bias, less HARKing, less p-hacking, less data dredging, less drawer effect). This would also lead to a better distinction between predictions and post-predictions and a better understanding of what science really is.
PBR: What has surprised you most?
Since founding PCI, what has most surprised you: the enthusiasm of researchers, the resistance from institutions, the difficulty of changing assessment culture, or something else entirely?
PCI: What surprised us most was the gap between the enthusiasm for the idea of PCI and the willingness to submit papers to it, especially in the early years before we created Peer Community Journal.
Researchers were very willing to contribute as reviewers and recommenders, but much more hesitant to submit their own work. This showed us just how strongly researchers’ careers are still tied to the prestige of established journals and, in particular, to impact-factor-based evaluation.
Another surprise was the difference between librarians and researchers. Librarians immediately understood the value of PCI because they are very aware of the costs of scholarly publishing. Researchers, on the other hand, often don’t see these costs directly because they are covered by grants or institutional agreements.
At the same time, we have received remarkable support from institutions, including many prestigious universities and research organizations. The paradox is that many of these institutions still rely, directly or indirectly, on journal prestige when evaluating researchers because they remain influenced by international rankings.
To us, this remains the main obstacle to changing scholarly publishing. The good news is that things are slowly evolving, particularly in Europe, where research assessment is increasingly focusing on the quality of the work itself rather than the prestige of the journal in which it appears.
--
Denis and Thomas, I am truly grateful to you both for taking the time to answer these questions so fully.
There is much here to think about: the real cost of publishing, the role of journal prestige, the difficulty of changing researcher behaviour, and the possibilities opened by diamond open access, transparent review and community ownership of scholarly evaluation. PCI is important because it shows that another model is not only imaginable, but already operating.
For those of us concerned with academic renewal, there are serious lessons here. Publishing reform will not come from criticising commercial publishers, however justified that criticism may be. It will require researchers, institutions and funders to support credible alternatives, use them, recognise them, and stop treating journal prestige as a proxy for scientific quality.
PCI has done the hard work of building such an alternative. The wider research community would do well to get behind it and help make a difference.