Serfs, Predators, and Robots: The Past and Future of Scientific Publishing

Published in Cancer

Share this post

Choose a social network to share with, or copy the shortened URL to share elsewhere

This is a representation of how your post may appear on social media. The actual post will vary between social networks

Academic scientists constantly receive emails to solicit submission of manuscripts from some unknown journals or are invited to be a speaker at newly invented conferences in nice tourism locations. Many of these notifications contain laughable flaws: incorrect spelling of the receiver’s names and/or titles, irrelevant field of research, wrongly cited papers of the subject, and grammatical errors. The so-called editor’s name looks so fake, sometimes only a given name is provided. An even more humorous aspect is asking the recipient to respond in a minimal number of days.

The senders of these solicitations are infamous predatory journals and predatory meetings. The fraudulent nature of these invitation messages is so obvious that it is amazing these businesses are still functional. In fact, they not only work, but they also proliferate and flourish.

So, as academic researchers whose major job is to publish papers, how did we get here and where do we go from here? To answer these questions, I will address the recent history of scientific publication and the future of scientific papers.


The building of serfdom

In the late 1990s, the biomedical research community received the biggest gift ever created—the U.S. Congress doubled the National Institutes of Health (NIH) budget! That was the largest historic leap of faith for science since the 1950s, back when “the manager placed a bag of money in front of the lab door and left quietly.” The next day, when the celebration parties died down, all the research institutes geared up to get money and expand. Suddenly, new faculty members were being recruited from everywhere, and the size of laboratories increased significantly. In fact, I did my Ph.D. studies in a lab with more than 50 people! These increases were not just restricted to the United States; many countries followed our example to increase the national budget of scientific research. Research institutes expanded, and the number of faculty members surged all around the world.

Because the salary of faculty can be fully paid by research grants, institutes found that with more faculty, more profit could be made from overhead charges. Faculty and trainees work in an environment owned by institutes. they have to generate income for themselves, contribute to the revenue of the institutes, and pay for the maintenance of the workplace. This is similar to serfs, in the Middle Ages, by definition. Many new research institutes were founded primarily driven by the goal of seeking federal grant money. The number of faculty members kept increasing and in a few years, the situation began to look more like, “The tragedy of the commons”—the famous case of unlimited access to grazing pastures for all sheep herders in ecology textbook. As the number of grant applicants shot up, a universally acceptable filter was required to comb through the competitors. Meanwhile, scientific publication had reached a fully mature stage and become an industry. Scientific editors took the interest of general readers into consideration when justifying acceptance. Peer review was introduced early in the 20th century and was gradually refined. Because research papers were typed on typewriters until the 1980s, copy editors played a key role in quality control by stringent examination of writing and formats. The quality of research papers appeared to reach a climax not seen before. Publication of research in peer-reviewed scientific journals became an obvious goal for scientists, for its perceived objectivity.

The resulting aphorism, “publish or perish,” is nothing new. One key aspect that separates this overcrowded new era from previous academic research situations is the importance of the quantification of the criteria. The impact factor of scientific journals, which is a summary statistic based on citations of papers published in a journal, became the absolute indicator of research performance. Building on this concept, the career of an academic researcher can be compared by using any kind of creative, quantitative index, such as the number of papers published in high-impact factor journals, the accumulated number of citations, h-index, and the total number of published papers over a given period.

Publication, and of course success at obtaining a research grant support, became an unequivocal and absolute measure of academic achievement, leading to a surge in the number of published papers. While welcoming this, those powerhouses of scientific journals further developed a two-pronged strategy to maximize profit. On one hand, they reduced the number of published papers in each issue of their flagship journals, so as to increase the average citation number per paper annually (which is linearly correlated with the impact factor). On the other hand, they expanded the brand-name series journals to accept as many submissions as they could, encouraging researchers to keep publishing. As a consequence, the total number of publications grew exponentially, while the impact factor of “top-tier” journals also increased beyond the contemporary imagination. Before the era of online publication, some journals also enhanced their profit by charging exorbitant for color figure printing and eventually page printing fees. They became the precedents of the publication fee introduced later.

At this point the strategy of research has a very clear goal: getting as many publications in journals exhibiting higher impact factors as possible; the more and higher, the better it is for the career. Over time, the top priority of research became to publish a paper, and science, while still important, was secondary. In this way, the measure becomes the target. The studies are directed to fulfill the expectation of the reviewers. Paper writing evolved to be more of an art that intends to convince the readers instead of showing the discovery. The editors of top-tier journals chase the hottest topics in the fields and brightest stars in the crowds. The New York Times' slogan is, “All the News That's Fit to Print,” and now the slogan in the academic community is, “All The Papers That Are Newsworthy.” Meanwhile, the lower-tier journals were flooded with submitted manuscripts. Publishing or perishing; survival is everyone's top priority.

The rent-seeking system uses a single measure to drive academic research more and more like a serfdom. People started to consider this system unsustainable. Someone thought to free the serfs.


The floodgate has opened

Traditionally, since the 17th century, scientific journals have been published by academic organizations whose elected/selected members serve as editors. It is still the case for many scientific journals, sometimes called “(academic) society journals,” today. In the 19th century, commercial publishing houses began to publish scientific magazines. Some of them evolved overtime to become scientific journals, too. However, instead of researchers themselves, the companies hire the professional publishers and editors to run it (referred to as “commercial [scientific] journals” from here). At the beginning, scientific journals were just one of the products of these companies, among many kinds of books and magazines. Through trading and merger, nowadays some of the commercial scientific publishers have become gigantic, multi-national, and powerful business entities for profit.  

Especially for commercial journals, scientific journals profit mostly from institutional subscription, for which only the affiliated members have access to the articles (evolved from the bound journals in the libraries to the password of the website!). Since reading and publishing high-impact articles are essential for both research and funding applications, the institutes will have to subscribe to such journals. The commercial scientific journal publishers, therefore, have every motive to promote the impact factors of their products. On the other end of the pipeline, as long as the number of publications and the impact factors keep dictating the measure of academic performance, the journals will have an endless supply of the good articles to boost their impact factors. The producers are customers at the same time so the profit can be generated in a positive feedback manner. However, it should be noted that the increase in the NIH budget alluded to above did not directly lead to institutions having more money to pay for a journal subscription. This required a high-level decision, typically taken by the Vice President for Research and the Head Librarian, at each institution to use some of the “indirect cost” money from grants to increase the budget of the scientific library and thereby pay for more subscriptions.

Maybe we can liberate the poor researchers by abolishing journal subscriptions? If everyone is free to read any article they want, the article can have an impact on all the researchers directly, not only through the journals by subscription. By this way, the impact factor of the journals will lose its impact. Open-access journals, most prominently the Public Library of Science (PLoS) were created with this great vision and with great fanfare. However, without charging a subscription fee, how could open-access publication be run? The answer: the authors pay for the publication of their own article. Economically, this model would also be a necessary change for the scientific publishers at that time because of the limited library budgets explained in the preceding paragraph. Thus, the pre-1990s economic model that subscription and advertising fees covered most of a journal’s budget was economically unsustainable. This imbalance in budgets effectively forced the “author pays” model. Moreover, NIH’s rollout of PubMed Central and a similar system set up by Wellcome Trust further was forcing publishers to shift to this model by undercutting alternative sources of revenue.

This was not unprecedented. In the early 2000s, many scientific journals already charge the authors a “printing fee,” or page charge. I still remember when my very first paper was published, it was noted at the bottom of the first page that, since the authors are charged with page fee, this article is deemed as an “advertisement” in compliance with some law. I was both happy and sad to receive the reprint of my paper—my research was a business that made money for others! Open-access journals expanded such charges to build a whole business model on publication fees paid by authors, eliminating subscription. Some other innovations were introduced by these new open-access journals around that time; for example, recruiting authors to be associate editors when they submitted manuscripts. Importantly, the policy of those early open-access journals was that reviewers were judges, not supervisors: they should only evaluate the presented results and the authors’ conclusions, not request new studies. Importantly, a key justification for the page charges in the 20th century was that page charges paid the salaries of the copy editors, but this arrangement got destabilized soon—stay tuned on this topic in Episode 3!

Open-access publishing was an audacious experiment to break the spell on the academic researchers. The open-access journals were championed by some very prominent scientists, including Harold Varmus, who regarded the “author pays” model as politically correct in going against for-profit publisher giants. Notably, the spectrum of papers in these founding open-access journals, especially PLoS ONE, fundamentally changed the perception for what is publishable. How does it work? Well, like most reforms in history, it did not develop as the designers expected exactly. It was intended to ditch the impact factor; however, in the next year after their launching, people were calculating the impact factors of these open-access journals. This reflects the stubbornness of human beings’ mind-set. The publication fee soon became a hot topic of debate; some believed that it aggravated the inequality in academic research between rich and poor institutes, or even countries. Moreover, the major publishers reacted quickly to launch a whole series of open-access journals to fight for this new territory. Unfortunately, they gave the open-access journals a “second-tier” look.

Ultimately, the open-access journals did not liberate researchers from the tyranny of impact factors but created a new market. Moreover, it opened Pandora's box.


The rise of the predators

Since the 1980s, the advances of information technology have been quietly re-shaping the scientific publication industry. First, the development of automated typesetting systems, such as SCRIBE, troff, TeX, LaTeX, eliminated one of the main functions of journal editorial offices. The copy editors had to justify their salaries by more stringent text screening, leading to a persistently adversarial relationship between copy editors and authors. Some newcomers in the publishing business had the brilliant economic insight that aggressive copy editors had become a liability, not an asset, because they slow down the publication process and annoy authors. Second, in the early 2000s, publishers were quick to roll out manuscript management systems, but slow to shift to online-only publishing. Therefore, it became much easier and faster to submit manuscripts, but the number of printed pages per journal per year was still constrained. The failure of the traditional publishers to go online-only led initially to the proliferation of new online-only publishers, starting with BiomedCentral in 2000. One of the many ironies of this story is that BiomedCentral is now owned by Springer/Nature.

All these changes prepared the niche for the new species.

The urban legend goes like this, in the late 1990s, a foreign postdoc in New York was disappointed with his future. His research was stuck and publication in top-tier journals was unlikely. His chance to find a faculty job either in the United States or in his home country was close to none, and he felt desperate.

One day he saw a newly launched open-access journal. He was awestruck. No paper printing, no sales of subscriptions to the big institutes. All it took was a website and an online payment system, and someone else would cover both production and cost. He could not have thought of an easier business: build a website for a research journal and serve as the publisher, editor-in-chief, and reviewer all in one. The website was launched, and the money poured in. It is rumored that the postdoc earned the first bucket of gold in his career, and I can imagine that he is enjoying life on a nice beach somewhere now.

One of the earliest open-access journals was Nature and Science. Being able to tell colleagues, “I published in Nature and Science,” had to feel like a huge accomplishment This journal is still around today.

Whether or not this is what really happened, it has become the standard business model for all such pay-for-publication “journals.” In many cases, it is felt almost like a one-person business where all the “unnecessary” operation costs, or any measure that slowed down the turnover, were cut, such as scientific and copy editors, reviewers, and the review process. They send out emails to solicit submissions, like predatory animals hunting for their prey, earning them the inglorious designation: predatory journals. They also inspired the emergence of predatory conferences.

Many emails soliciting manuscript submission are generated automatically, almost at the high school level of amateurism. For example, in those emails, my name has been written as “Day CP” (obviously a direct copy from PubMed), misspelled, or simply left blank. In many cases, irrelevant papers, or even corrigendum, were cited in the soliciting message. The names of the “editors” were also collectively funny. In many cases they look simply made up. The whole format is so similar to telemarketing scammers that one would suspect it is a bad joke.

How is it possible that someone would still submit manuscripts to these “journals,” it is like prey walking into the mouth of the predator? It is because there is a demand and, thus, a market. In the 1990s, many countries have followed the example of the United States and increased their investment in scientific research. Also, like the United States, many governments evaluate the performance of research using publication as a measure, but with a twist: they request that researchers publish their papers in so-called “internationally recognized scientific journals,” meaning the papers are written in English, the journals are indexed by PubMed, and the publishers located in the United States or Europe. Many predatory publishers register their addresses in New York or Basel, but fully operate somewhere else. After all, all it takes to be a “scientific publisher” is a website in English and, of course, an online payment system. To meet the demand of the research industry, the predators proliferate.

As the research community began to realize the damage that predatory journals can cause, some efforts were made to control them. In 2008, Jeffrey Beall, a librarian at the University of Colorado, started to receive numerous requests from dubious journals to serve on their editorial boards. He immediately became fascinated because most of the emails contained numerous grammatical errors and investigated these supposed journals. He coined the term “predatory journals” and later developed a list of potential, possible, or probable predatory scholarly open-access publishers, which is now known as “Beall's list.” Its length grew from 18 publishers in 2011 to 923 in 2016. Beall and his university received multiple lawsuit threats, forcing him to remove the list from his blog in 2017. Nevertheless, he has since expanded his investigation to “predatory conferences.”

The journal Science also launched a sting operation. The editors submitted fake manuscripts with ridiculous content to the journals on Beall's list, 80% of which accepted the fake submissions. The accuracy of Beall's list was debated following this result, but how discerning the peer review process was for those journals that rejected the manuscript is not known.

Since then, many predatory journals have been exposed and discredited. However, the demand for research publications still exists, and all the predators need to do is find a new resource and niche.


The evolution of the beasts

The central theme of predatory journals is to lower operation costs or eliminate them altogether. All it takes is a website (and for sure a payment system) and the publisher can sit back, wait for manuscripts to pour in, and the authors pay for everything such as, editing and publication. The lack of any necessary but cost-taking procedures in regular scientific publication, including editorial processing and peer review, deems them illegitimate.

As the “legitimate” journals chase high-impact factors by limiting the number of papers published in each issue, many researchers are left needing to publish their work in a source that will be taken seriously, not viewed as some phony laughingstock material. The market is huge and open, just waiting for someone to take it.

How could one have all the legitimate procedures and no cost? Some predatory publishers found an abundant, almost endless supply of resources: the many researchers in the world who are eager to get credit and recognition. The titles of manuscript reviewer and journal editor would be wonderful decorations on a CV, even bestow some power. In a serfdom where land is scarce, anything can help. The expertise, time, and enthusiasm are there to tap.

A few predatory publishers figured this out. Rather than just using these researchers as a façade, why not ask them to do the real job for free? They hired editorial managers to run the process but keep all the processes on time. It is the “light” version of traditional journals, run on the basic management cost. Though some garbage is filtered out so not all submissions are published, the abundant supply of manuscripts still makes it very profitable. Moreover, since these are real results, these journals even gain decent impact factors. The predators evolved to learn not to deplete their resource. They filled this new niche and managed to make it sustainable, just paying some basic costs. One of them knows how important formality is for academia. It decorates the papers with delicate page designs and pretty illustrations, which work very well to attract readers and even citations.

This is a great business model that will not be missed by the traditional powerhouses of scientific publishers. They copied the strategy, tapping the researchers to be reviewers or even editors for free, to publish their own “light-version” series journals. Nowadays you get emails every week to invite you to be a guest editor for a special issue or a reviewer. In a way, this is a win-win situation: the researchers can publish their studies while getting credit as editors/reviewers, the publishers make a profit, and together they effectively squeeze the space of the “conventional” predatory journals in the market. When you can publish in real journals, why go to a fake one?

However, looking at this immense market, some of these publishers cannot help watering down the already very light version to either further reduce costs or to take more of a market share. One example is a publisher that keeps expanding its journal portfolio. Presently it publishes more than 400 journals and is still expanding. In another example, a publisher’s one-person “editorial office” manages more than 200 journals. That person’s only job is to send the manuscript to the guest editors and invited reviewers, and ask them to do all the jobs, including checking spelling or the number of figures. I have found that the manuscripts sent by such publishers all looked the same, including the rationale, method, and analyses. And they were all from the same country. They were obviously generated by paper mills. I emailed the so-called “editor” about this situation and got a reply that it was not his job to care about this. Obviously, that would be my job. Such strategies make these publishers look like the “cheaters” in an ecosystem.

While top-tier journals and impact factors still rule academia, the “light-version” publishers began to change the perception of scientific papers. Researchers publish papers in these journals more and more, especially those from the younger generation. As readers find more interesting studies from these fringe journals, they read, publish, and cite them more frequently. Recall from the introduction that a key driver in all these changes is the desire of traditional journals to have a high-impact factor. The official impact factor of a journal has traditionally been determined by the Science Citation Index, which is now called Web of Science and owned by Clarivate. But simultaneous with all the above developments in publishing the authoritative status of licensed Science Citation Index, which uses only peer-reviewed sources of citations has been threatened and likely surpassed. The newer more widely used authority of citation counts is Google Scholar for the obvious reason that Google Scholar is free of charge. However, Google Scholar also differs from the Science Citation Index in that Google Scholar takes citations from all online sources, including nonrefereed sources, which indirectly damage the reputation of peer-reviewed sources. This shift to include more sources of citations also resulted in the inflation of h-indices. Therefore, papers published by the “light-version” journals proliferate in citations. Quantity accumulates to drive the change in perception. The impact factor starts to not look so impactful.


Liberator ex machina

Another branch of the open access movement began to spread when the fringe journals were trying to twist into the mainstream: preprints. In fact, it was the origin. Around 1990, Joanne Cohn began emailing physics preprints to colleagues as TeX files, but the number of papers being sent soon filled mailboxes to capacity. Paul Ginsparg recognized the need for a central storage location and created a central repository mailbox stored at the Los Alamos National Laboratory in 1991. In 2001, Ginsparg got support from Cornell University, moved the repository there, and changed the name of the repository to, which stored preprints and postprints (known as e-prints) in the quantitative studies (e.g., physics, mathematics, computer science, etc.). Posting an article requires approval after moderation, but not peer review. Submission to arXiv has continued to grow quickly since its launch. Anyone can read and share the manuscripts without restriction. Among physicists, there is a common refrain: “If it’s not on arXiv, it doesn’t exist.” Its success changed the norm of scientific publication or marked the return to the old tradition of scientific communication and inspired open-access journals later. In January 2022, arXiv began assigning digital object identifiers (DOIs) to articles.

At the peak of predatory journals, in 2013, John Inglis and Richard Sever founded bioRxiv, which is hosted by the Cold Spring Harbor Laboratory, for preprints of biomedical sciences. Like arXiv, the manuscripts are posted after moderation, and there is no restriction for reading and sharing them; readers can also post comments under each manuscript. Researchers soon found that preprints in bioRxiv can function as proof of original studies that prevent others from scooping ideas. Moreover, each preprint is given a DOI and can be cited. It can show the readiness of a research project when applying for funding. The number of preprints submitted and posted has grown exponentially since the launch of bioRxiv.

About two-thirds of the preprints in bioRxiv were eventually published in conventional peer-review journals. Also, after posting, the quality of the studies in the preprints can be reflected by the number of citations, tweets, and retweets. Readers can also decide if the data in the preprints are reliable by comparing them with similar studies searched in Google Scholar. Recently, ASAPBio and EMBO, in collaboration with bioRxiv, launched a service for forwarding the preprints to be peer reviewed. Upon receiving the request, the editorial team will look for experts to volunteer to review the manuscript. The review focuses on the quality of the study, not the so-called “interest of general readers.” Therefore, the process can be completed in 5 weeks.

At this point, what are the differences between the repository of peer-reviewed preprints and current biomedical journals? The biggest difference is that, for the former, the readers can decide by themselves if an article is good enough to be downloaded and read; for the latter, the editors make decisions on what articles to publish. The former is free of charge and the latter charges thousands of dollars. Other than those, both post peer-reviewed articles that can get cited and indexed by Google Scholar and PubMed (for now). The boundary between preprints and journal articles has begun to blur, not to mention the high retraction rate of articles from top-tier journals.

Imagine a robot is created to perform the basic QC check, categorization, and identification of potential reviewers as a manuscript comes in. When the paper is posted together with the review comments, the robot will also analyze if the results pass statistical tests or were consistent with historical data. You know your study will certainly be posted/published and assessed in a reasonable timeframe, without worrying about the cost. When you begin to plan a study, the robot will identify the most relevant studies from the public repository and organize the information for you. In this scenario, how much will you still care about whether a journal has a high-impact factor?

What can the traditional scientific journals do after robots take over manuscript publication? In fact, one of the biggest reasons why we fight so hard to publish in top-tier journals is to advertise the significance and impact of our research. I believe the answer lies in a quote a very wise scientist once gave me, “All papers are newspapers.” There you go! The journals can return to their origin—journalism for science. The career editorial teams can still write editorial opinions and comments and do investigations. They can still generate impact by selecting and highlighting the manuscripts that they find important or interesting.

If I had written this article a year ago, it would be a sci-fi story. Up to where we are now—with the advances of AI and human-machine interfaces like ChatGPT—the scenario described here would be a coming reality, not a fantasy anymore. This might be the beginning of liberating the serfs.



The author would like to thank Sarah Spaeh, Ph.D. for editorial assistance and Alejandro Schaffer, Ph.D. for his helpful suggestions and discussions. Additionally, the author thanks Yolanda L. Jones, NIH Library for editing assistance.

Update: some interesting articles

How big is science’s fake-paper problem?

ChatGPT use shows that the grant-application system is broken

Scientists prefer feedback from ChatGPT to judgement by peers

The strain on scientific publishing

Scientists paid large publishers over $1 billion in four years to have their studies published with open access

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in