Preprints

Many months may elapse between the submission of a manuscript and its publication in a journal. According to one study (Huisman & Smits, 2017), the average duration of the peer-review process at scholarly journals is 17 weeks (Huisman & Smits, 2017). In addition, time is needed for formatting, copy-editing, and other work steps. To expedite the dissemination of research findings, preprints are shared in many disciplines. Preprints accelerate the sharing of research findings and enable early feedback (Blatch-Jones et al., 2023; Smart, 2022). 

In addition to the quicker and more effective dissemination of research findings, the public release of preprints may also create challenges for the peer-review process. Especially in the case of blind peer review, the public availability of a manuscript may make anonymity more difficult (Rastogi et al., 2022; Smart, 2022). As possible solutions, publication venues may require alternative titles and slightly modified abstracts during the peer-review process. This makes it more difficult for the reviewers to clearly identify the corresponding work online. A further possibility is to make anonymous preprints available to the public via platforms such as OpenReview. By so doing, authors can make their research accessible during the review process and can associate the work with their names once the blind peer review has been completed (Rastogi et al., 2022). 

However, cultures differ greatly across disciplines. In physics, preprints are an indispensable communication tool – depending on the subfield, preliminary versions of over 90% of all journal articles are made available to the public on the preprint server arXiv prior to publication (Gentil-Beccot et al., 2009). Especially since the COVID-19 pandemic in 2020, preprints have also become increasingly important in other disciplines, for example, mathematics, computer science, biology, medicine, and, to a certain extent, economics and the social sciences. However, it should be borne in mind that not all journals accept manuscripts that have already been made publicly available as a preprint. 

As a rule, preprints have not yet undergone any scholarly quality assurance. Substantively, therefore, they may differ considerably from the published version or may not even be accepted for publication. While the scientific community is aware of this, problems may arise if, for example, the media pick up findings from preprints and disseminate them to a broad audience without indicating that they are preliminary and have not been peer reviewed. To enable non-scientific readers to better understand the nature of preprints, medical preprint servers display caution statements. 

As early as the 1960s, there was a system in biology whereby reports of current research findings – especially in the form of preprints – were reproduced and circulated by postal mail by a central registry (Cobb, 2017). A similar system also existed in the field of high-energy physics (Gentil-Beccot et al., 2009). With the advent of the Internet, the sharing of manuscripts became much easier, as they could be circulated via email. In 1991, Paul Ginsparg founded arXiv at Los Alamos National Laboratory as a server for preprints from high-energy physics (Ginsparg, 2017). The manuscripts were thus freely available to all interested parties. The scope of arXiv was later extended to include other areas of physics, as well as mathematics, computer science, and neighbouring fields such as quantitative finance and quantitative biology. From 2001 to 2026, arXiv was operated by Cornell University; since 1 July 2026, it has been an independent non-profit organisation. 

It was only after the turn of the century that other fields followed arXiv’s lead. For example, in 2013, bioRxiv was launched as a preprint server for biology, thereby making preprint sharing more popular in the life sciences. The initiative ASAPbio advocates the use of preprints in biology. Boosted by the success of the Open Science Framework (OSF) – an open source infrastructure that, inter alia, enables preprints to be uploaded and searched for – there has been a veritable boom in preprint servers for the most diverse subject areas since 2016. Nonetheless, the acceptance of preprints and the willingness to share them varies greatly across disciplines. During the COVID-19 pandemic, preprints played an important role in the rapid dissemination, checking, and correction of research findings on the SARS-CoV-2 virus (Fraser et al., 2021; Gianola et al., 2020). 

Preprint servers enable the publication of overlay journals (Rousi & Laakso, 2024) – that is, online journals that use the infrastructure of a preprint platform, especially arXiv, for submissions and publication. Authors upload their paper to arXiv and submit a link to the arXiv preprint to the journal in question, which then organises the peer-review process. If the paper is accepted, it is included in the journal – that is, linked from the journal website and assigned a number, a DOI, and possibly a comment from the editor. In this way, the advantages of arXiv (cost-effective, freely accessible, well-known infrastructure) are combined with those of a journal (peer review, reputation, possibly an Impact Factor). Examples of successful overlay journals are SIGMA (Symmetry, Integrability and Geometry: Methods and Applications), Logical Methods in Computer Science (LMCS), Discrete Analysis, and Quantum.

The overlay journal movement is still in an early stage. Compared with the total number of open access journals, the current number of overlay journals is extremely small. However, studies (e.g. Rousi & Laakso, 2024) have shown that new overlay journals have been actively established in recent years. 

Although most overlay journals are concentrated in the fields of computer science, mathematics, and the natural sciences, some recently founded overlay journals are in the fields of medicine or the biological sciences or are multidisciplinary (Rousi & Laakso, 2024). As a rule, overlay journals are published by groups of academics rather than formal organisations (Rousi & Laakso, 2024). 

Preprints not only expedite the dissemination of scientific findings but also increase visibility, accessibility, peer feedback, and transparency in the publication process (Blatch-Jones et al., 2023; Smart, 2022). At the same time, they create new opportunities for open peer review and for a more community-driven discussion of scientific quality (Avissar-Whiting et al., 2024). Preprints are therefore not only a preliminary step on the way to journal publication but also an independent and increasingly important element of open scholarly communication.