We live in a time when there are both reasons to celebrate, and reasons to deeply worry, about scientific publishing. On the one hand, scientific publishing is growing rapidly, driven by the growth of the higher education sector, research expenditure, and internationalization. Scientific publishing is also becoming more globally balanced, with the rapid growth of scientific activity in middle-income economies (NCSES, 2021). On the other hand, there are concerns that the current system of scientific publishing is in crisis, with concerns raised about the quality of peer review and publication inflation (see Flaherty, 2022 and Whitehouse, 2015), and the cost of the current system, whose financial benefits appear to accrue dis-proportionally to academic publishers (see Hagve, 2020 or Puehringer et al, 2021). So, how did we get here?
The scientific publishing system we have today emerged in Europe during the seventeenth century, and used printed periodicals as its medium. This system lent itself to a subscription-based economic model of publishing, whereby periodicals were purchased mainly by the libraries of universities and other research institutions.
The internet changed the scientific publishing system by shifting communication and distribution online. Publishers offered access to their journals through online subscriptions, while the internet also enabled an “open access” publishing model, whereby publishers collect an upfront fee, while providing free access to scientific papers via a website.
However, both the subscription and the open access systems, do not appear to address the most important problem in scientific publishing, which are (1) a reliance on unpaid volunteers for review and most editorial work, both of which play a critical role in scientific quality control, (2) the distribution system of scientific publications which is controlled by publishers and not by other stakeholders, and (3) copyright, which is usually held by publishers, while most authors receive no compensation for their work. These three issues are expanded upon below.
Volunteer-based Review and Editorial Work
Scientific publishing relies deeply on peer review, a critical service that is usually provided by anonymous volunteers. Peer review is not only a system of scientific quality control, but it is also an opportunity for learning by the authors and reviewers, and the process often enhances the quality of the manuscript.
As academic competition grows, the incentives for reviewing have declined. Academics themselves face pressure to spend their time doing research, rather than to review the work of others. The hours spent are also difficult to justify to funders: no direct research output is created, it does not directly benefit student learning, it also does not directly enhance the reputation of the university. The only compensation reviewers receive is nominal recognition, for instance through membership of an editorial board, or more recently, through score-keeping via Publons.
The current system therefore leads to sub-optimal outcomes: fewer and lower quality reviews, both in terms of the level of detail and time spent by reviewers, as well as the use of reviewers with limited expertise and experience. Malevolent authors have also been successful at gaming the volunteer-based peer review system in more than one instance (example, example).
The fact that the critically important core function of peer review in scientific publishing relies on unpaid volunteers is a clear weakness of the current system.
Vulnerable Distribution System
Scientific work is typically distributed via the website of the publisher, and while convenient, this also makes it vulnerable to loss or interruption. Websites can be easily censored, both by the publishers themselves, who may respond to pressure from governments or libel suits, and by internet service providers, who may be ordered to block content. Furthermore, publishers can simply close down, making their scientific content disappear from the web.
There are ways in which this scientific content can be preserved despite these vulnerabilities, for example through copies kept in repositories, or on personal websites. However, it is not always clear if these copies are the genuine final product, and they may be more difficult to find than the original published article.
Yet the internet offers a range of decentralized content distribution technologies, including BitTorrent and IPFS, which can help preserve and distribute scientific content. While decentralization does not guarantee content persistence, it becomes much more likely, especially if the organizations who spend significant amounts on academic journal subscriptions, also provide some resources to mirror content, similar to how open source software projects like R and Python are supported by large institutions.
Copyright and Governance
Publishers typically retain ownership of scientific articles and journals (including their title), which hinders the decentralized distribution of content, and gives publishers considerable economic power. An editorial team cannot change publishers, or change the economic or governance model of a journal without the publisher’s consent, because they are not the owners. While a publisher-led journal governance model has its strengths, it may not be the only possible model for evaluating and distributing novel scientific output.