As academics transition to digital journal publishing, the Open Access (OA) movement, which aims to make research freely available at or soon after publication, is picking up record speed. According to a recent Simba report titled Open Access Journal Publishing 2016-2020, “the number of OA research articles published annually is growing at double the rate of the complete spectrum of research articles.” The report also found that about a third of all research articles currently published are OA, when factoring in those with complete embargo periods.
The rise in OA publishing comes as a result of scholars banding together to reveal the mounting cost of research access for institutions and individuals, such as those behind the Cost of Knowledge Project who started a boycott of Elsevier, which now has profit margins exceeding 30%. At the same time governments and funding bodies have introduced new OA mandates and calls for early sharing of research. Such as the Wellcome Trust, Bill & Melinda Gates Foundation, and others who this February urged journals and research funding agencies to sign a pledge to make all papers related to the Zika virus OA in the wake of the public health crisis.
Yet, despite the scholarly community embracing OA publishing, the Authorea team reveals that 65 of the most cited papers in the world are still behind often hefty paywalls, inaccessible to those who can’t afford article fees or journal subscriptions. Today this pattern continues with more groundbreaking articles being published in journals with prohibitive paywalls. As the drumbeat for OA gets louder, it’s apparent that researchers and journal publishers must come together to find sustainable ways to make such research freely accessible.
Among promising OA solutions is preprint servers, or online databases of manuscripts posted by scholars prior to formal publication.
In this guide we’ll look at the place of preprints in the digital publishing landscape for journals and scholars, including:
How preprints are helping scholars and journals make research more open
Challenges to be addressed when publishing via preprints
New publishing models journals are pioneering using preprint servers
Let’s get started!
Preprints and the modern publishing landscape
Researchers utilize the most advanced tools in the world to perform their experiments. From massively parallel computing clusters approaching exoscale computation power to high-powered microscopes redefining the limits of resolution, we're in age where the tools we use to perform research are truly awesome. However, such innovations have fallen short when it comes to how researchers write or ultimately publish their work. In short, scientists are doing 21st century research, writing it on 20th century tools, and packaging it in 17th century formats.
The origin of scholarly journals: 300 years with little change
The first scholarly article, published in 1655 in Philosophical Transactions, was an important step for researchers towards formalizing and preserving the scholarly record. Recognizing the importance of research communication, the opening editorial stated:
"Whereas there is nothing more necessary for promoting the improvement of Philosophical Matters, than the communication of such."
Over the ensuing centuries new publications from The Lancet in 1823 to Virchows Archiv in 1847 were launched (fun fact: Thomas Wakely launched The Lancet at age 27 and Rudolf Virchow launched Virchow's Archiv at the age of 26). Considered radical in their time, these journals and others soon grew in prominence becoming the standard method for communicating research. Many of which are now household names and contain some of the world’s most important findings.
The increasing volume of academic publications brought about a formalized peer review system in the early 19th century and with it an increase in technological capabilities over time, like structured typesetting, which were largely born out of a desire to share research better. In the 1960’s the way scholars consumed and disseminated their research began to change... at least for some researchers.
In the 1960’s, while the majority of researchers communicated via journals exclusively, a subset of researchers began to utilize new forms of communication. Physicists began to share their paper manuscripts directly with each other prior to them being peer reviewed or published in a journal. These shared manuscripts become known as “preprints.” Databases in the US and in Germany housed the bibliographic information of preprints and facilitated their sharing, which eventually became so widespread it became onerous and even unfeasible to manage for those without adequate resources. Indeed, the library at the Stanford Linear Accelerator Center (SLAC), which housed preprints, was receiving thousands of preprints per year at one point, a large amount of information at the time to be organized and collected.
Preprints and the web
To deal with the growing amount of bibliographic information accumulating at SLAC a computer database was developed, termed the Stanford Physics Information Retrieval System (SPIRES). SPIRES not only organized bibliographic information, creating a standard used across universities, but also allowed for researchers to email the database and request that a list of preprints be sent to them. Since papers could not be emailed at the time, the system relied on snail mail.
While SPIRES greatly improved the flow of information, it often still took weeks for preprints to reach their requester’s mailbox. A typesetting system called .tex introduced in the late 70’s early ‘80s soon changed this by allowing researchers to write their documents in a specified manner that could be emailed and downloaded and compiled without the need for physical mail.
Soon researchers, primarily physicists, were emailing and downloading .tex files at great rates hastening the process of research communication, which again created the problem of information overload. To manage this process, in March 1989 Tim Berners-Lee wrote a paper modestly entitled, "Information Management: A Proposal." Soon after the World Wide Web, Berners-Lee’s own invention, and with it a central preprint repository, later to be named arXiv, were born.
What researchers are saying about preprints
The rise of open research sharing, the web, and a central preprint repository drastically changed how researchers communicate their research. However, these advances did not penetrate all research disciplines and to this day preprints have struggled to gain widespread adoption. Nevertheless, in recent years preprint usage has been on the rise and there are now preprint repositories for various disciplines including arXiv, for math, physics and other sciences; BioArXiv, for biology; SocArXiv, for social sciences; as well as discipline-agnostic repositories like Authorea.
"New technology does a better job of finding me stuff to read than the table of contents of any given journal (Pubchase, custom RSS feeds, Twitter recs from respected colleagues). Especially with the option of versioning articles, I think this system would capture much of the current value of peer review, be faster, and more sustainable."- Angela DePace, Assistant Professor, Department of Systems Biology, Harvard Medical School
"Posting preprints offers people the chance to be more thoroughly evaluated, which is especially beneficial for younger scientists."-James Fraser, Assistant Professor, UCSF
One limitation that has held back the widespread adoption of preprints is the confusion over whether you can still publish in a traditional journal after posting a preprint. This fear, despite the fact that most preprints on arXiv are ultimately published in journals, arises from the so-called Ingelfinger rule that limits duplicate publications as well as communication by some publishers that has cast an artificial uncertainty into the decision making process of researchers.
Nevertheless, there are clear rules for writing and posting of preprints and journals are quickly embracing preprints in wider disciplines. In fact, some journals have even instituted "preprint editors" to solicit preprints for journal publication off of repositories
Preprints reshaping the author experience
Preprint servers arose with the use of new technology, specifically authors writing their documents in .tex files. Today, much of the world is able to write online and even execute their own programs for data analyses. Despite this widespread adoption of technology, scientific papers remain primarily static objects (PDFs) and generally require typesetting to be done by journal publishers.
New collaborative writing tools, such as Authorea, simplify this process by allowing authors to produce manuscripts that are not limited to just text and image but also allow rich media such as datasets, software, source code, and videos. Tools like Authorea make it easy to write, cite, collaborate, host, and to preprint. By putting the control of typesetting in the hands of authors that write in markdown, latex or rich text control of publishing is shifted towards the research community, not just professional publishers.
5 Ways you can start using preprints to publish
Check preprint policies at journals to which you’re considering submitting
Email the editors of the journal you plan to submit to if you’re unsure about their preprint policies
Use collaborative writing tools to write and post preprints
Share your preprints with colleagues via social media and email to solicit feedback
Educate and encourage others to share their work openly
The role of preprints in academic journal publishing
In the digital scholarly communication landscape the primary value of preprints for researchers is that preprints allow them to disseminate their work faster and to a much wider audience than they could in traditional paywalled journals. A recent Editage Insights article titled, “The role of preprints in research dissemination“ details why many scholars are choosing to submit their work to preprints prior to formal journals. Among primary benefits are that preprints:
Give scholars working on time-sensitive projects a way communicate their research while undergoing peer review
Can serve as a way for scholars to establish priority over a particular research discovery or method
Ensure that an OA version of scholars’ work will be available regardless of where they formally publish
The increase in preprint use, along with scholars publicly coming together to endorse preprints, such as those at the 2016 ASAPbio meeting, is forcing journal publishers to address the place of preprints in the publishing landscape.
Preprints and Green OA
Despite “preprint” sounding like something reserved for pre-publication, “preprint” servers and published journals don’t have to be as separate as their names suggest. Preprints have actually been a driving force behind Green OA, in which scholars deposit their articles into an institutional repository or subject repository in order to make a version of their work OA. Depending on the requirements of the journal, Green OA preprints may be manuscripts or copies of final articles.
Among leaders of the Green OA movement is Stevan Harnad, Professor in the Department of Psychology at Université du Québec à Montréal and Affiliate Professor in Electronics and Computer Science at University of Southampton, UK.
In 1994 Harnad published the Subversive Proposal stating that researchers should self-archive their research articles to make them free for all online. Self-archiving later went on to become “Green OA” (and OA journals became “Gold OA”). Harnad also commissioned the first OA repository software, EPrints, in 1999. Since then, Green OA via repositories and preprint servers has become increasingly popular.
When asked how Green OA should work, Harnad says it’s simple.
“All researchers should self-archive their articles immediately upon acceptance for publication by depositing them in their institutional OA repository. And all institutions and funders should mandate this.”
How journals can facilitate Green OA via preprints
In order to use preprint servers to make research Green OA,scholars need journal support. “Open Access compliance: How publishers can reach the recommended standards,” a 2016 report released by Jisc, outlines some key ways publishers and journals can take action to help scholars make their research OA. Among steps Jisc outlines are:
Making clear OA policies easily accessible on journal websites
Ensuring embargo periods, if deemed necessary, meet funder requirements and are reasonable to the academic community
Applying Digital Object Identifiers (DOIs) to all articles, so scholars can link the different versions of their work
Adopting ORCID IDs, which can be used to populate institutional repositories and track OA compliance
Providing journal and article level OA licensing terms to clarify any differences at either level
As journal publishers continue to work out whether they’re ready to transition to an OA model, supporting Green OA by allowing authors to submit manuscripts they’ve posted to preprints is a great place to start. By considering manuscripts scholars have uploaded to preprint servers publishers can better meet institutional and funder OA mandates as well as scholar expectations. A key aim journal publishers must have to facilitate Green OA is to provide clear policies for preprints, so there is no confusion among scholars as to whether they can upload their manuscript to a preprint server prior to formal publication.
Importance of clear preprint policies
A recent article in the Journal of Librarianship and Information Science titled "What does ‘green’ open access mean? Tracking twelve years of changes to journal publisher self-archiving policies," traces the self-archiving policies over the last 12 years of the original 107 publishers listed on the SHERPA/RoMEO Publisher Policy Database. The study found that while the volume of publishers allowing self-archiving has increased by 12%, the volume of restrictions around self-archiving has increased by 119%.
Many preprint limitations enacted by publishers are not immediately apparent, making it vital for submitting authors and journal editors to carefully review publishers’ self-archiving policies to ensure they’re comfortable with the stances the journals they’re working with are taking. Confusing preprint policies can slow down Green OA and even deter scholars from submitting to preprint servers, due to unfounded fears of being penalized for publishing their paper in two places.
With OA publishing becoming an expectation among the academic community, today some journal editors are speaking out in instances of publishers increasing the cost of research access or enacting policies to limit OA options. In some cases where editors and publishers haven’t been able to work out differences editors have actually left the publisher to start OA friendly counterparts to old journals.
An early example is Michael Rosenzweig, Professor of Ecology and Evolutionary Biology at the University of Arizona, who in 1998 became one of the first editors-in-chief to leave his publisher when he led the editorial board of Evolutionary Ecology in declaring independence from the journal’s then publisher, International Thomson. Evolutionary Ecology’s editorial board left in protest of high subscription costs which, despite the editors’ best efforts to combat them, were rising at a rate of 19% a year. The editors started a new Green OA journal titled Evolutionary Ecology Research. Since then more journals have followed suit, including the editorial board of Elsevier journal Lingua, who also left in protest of prohibitive price hikes and started their own journal Glossa.
Peter Suber, director of the Harvard Office for Scholarly Communication, Senior Researcher at the Berkman Center, Senior Researcher at SPARC, and Research Professor of Philosophy at Earlham College, has been following the trajectory of editors leaving their publishers to start more OA friendly journals. Since the late ’80s he’s maintained the “Journal declarations of independence” list that he started on the Open Access Directory (OAD) website with OAD co-founder Robin Peek.
Suber believes that if publishers do not address prohibitive paywalls and OA policies it’s only a matter of time before more editors take a stand and launch solo journals on their own.
“Every time a journal declaration of independence takes place, frustrated editors at other journals will think about whether they should follow suit, because they will know it can be done,” said Suber.
Questions surrounding preprint publications
Preprints aren’t without questions to be worked out. Some publishers have been wary of fully endorsing preprints due to perceived challenges in the preprint model. Among those who’ve voiced uncertainty about preprints is Emilie Marcus, Editor-in-Chief of Cell and CEO of Cell Press.
Questions and uncertainties about preprints raised by Marcus include:
Are preprints by themselves valid publications and should preprints be citable?
If preprints are citable, do scholars have to read through all preprint literature in addition to published articles to account for preprints in their references?
Is there an optimal number of preprints per discipline and what should that be?
A concern among some scholars and publishers is that preprints don’t go through peer review, and post-publication vetting of preprints isn’t generally guaranteed or monitored. Though, pre-publication peer review is also an imperfect process with room for error. Uncertain manuscript vetting is a key concern among publishers when it comes to preprints, making some including Marcus, regard preprints as ways of accelerating the sharing of ideas but not a “publishing” solution. However, this isn’t the only way to view preprints. There are editors who are establishing new methods to link preprints and formal journals making research more open as a result.
Journals pioneering new preprint publishing models
A primary contention some publishers and editors have with preprints is that they lack the validation of formal journal articles. But what if preprint manuscripts were actually vetted and peer reviewed? Enter the vision of the overlay journal. Overlay journals are open access journals that perform the peer review function of a traditional journal but, rather than directly accepting manuscript submissions, pull in manuscripts from a preprint server.
Overlay journals started appearing in the 1990s beginning with Physical review D, the first partial overlay journal launched by Paul Ginsparg in 1996. From there more overlay journals followed. Generally, in the overlay model journal editors vet submissions, coordinate peer review, and then, rather than publishing accepted articles in an issue, they republish final versions to a preprint server along with a DOI. Overlay journals can be open access or use subscription models for final versions.
Some editors have been able to make their journals both free to read and free to publish in via the overlay model. Among examples are Peter Coles, Professor of Theoretical Astrophysics at Cardiff University, who founded the Open Journal of Astrophysics and Royal Society Research Professor and Fields Medalist Timothy Gowers who founded the new math journal Discrete Analysis. Both Coles’ and Gowers’ journals work in tandem with arXiv. The journals are formal publications, which can be abstracted, indexed, and accrue impact.
As a new overlay journal Discrete Analysis is taking steps to further develop the publishing model. Discrete Analysis stands out from other overlay journals because it has a formal publication website, which includes article categories visitors can choose from to access complete article listings and designated pages for each of its articles with an image and description. In this way Discrete Analysis offers readers a journal browsing experience on its website while still using arXiv to handle the publishing process.
Discrete Analysis manages peer review and publishes its journal via Scholastica. Using Scholastica’s arXiv integration authors are able to submit papers they already posted to the arXiv for consideration by Discrete Analysis editors. The journal’s editorial board then coordinates peer review for all submissions and accepted manuscripts are edited and re-uploaded to arXiv as final versions with DOIs. The journal has already published top research including editor Terence Tao’s article on crowdsourcing the solution to the Erdős discrepancy problem.
For Timothy Gowers and other overlay journal founders and supporters, the primary goal of the overlay model is to lower the barrier to publishing OA and speed up research dissemination.
“I want to be aggressively modern. I want to use the internet properly – when you’ve got something, you post it,” Gowers explained. “We’re not pretending to be a traditional journal, we’re something else.”
The overlay model presents a much leaner publishing approach that eliminates the costs of time and budget journals must usually allocate to “printing services,” such as typesetting, making the publishing process quicker and more affordable. It’s one of many new OA publishing models out there for journals to explore.
Research dissemination has not dramatically changed since the Royal Society published the first scholarly journal, Philosophical Transactions, in 1665. However, today’s digital publishing age is starting to transform the way scholars and journals approach research publication, spurring important questions about the traditional paywalled publishing model. With the shift towards digital publishing, OA, and new impact indicators, it’s becoming harder for publishers to justify the need for “printing” costs. Additionally, factors like a journal’s Impact Factor and the clout of its publisher are becoming less important to authors, making them more likely to consider alternatives to “band-name” paywalled journals. Rather, scholars, as well as universities, funding bodies, and government organizations, are encouraging researchers and journals to take steps to make content freely accessible. Preprint servers are one tool that is helping to facilitate Green OA via the archiving of preprints and even Gold (free to read) and Diamond (free to read and free to publish) OA via new overlay journal models that virtually eliminate publisher costs. Now's an exciting and significant time for the academic community to address the place of preprints in the publishing landscape. Tools for researchers and journal editors to do their work in a modern way are now at hand; it’s time to start utilizing them.