Academic researchers including graduate students, post-doctoral fellows, and university professors are intimately familiar with the concept of "publish or perish," a phrase which has been used since at least 1942. However, the peer review process—which provides quality control for published research—dates all the way back to 1665.
Since then, the internet has profoundly changed how researchers interact with each other and with the general public. As a result, some have argued that "publish or perish" should be replaced with "be visible or vanish", "promote or perish", "be discoverable or die", or "be cited or suffer". Let's explore whether "publish or perish" is still relevant today.
"Publish or perish" explained
The term "publish or perish" means that researchers are expected to produce a minimum volume of published scholarly work—typically in the form of peer-reviewed journal articles or scholarly books—in order to continue in their academic careers. A scholar's publication record plays a central role in determining whether a doctoral candidate can graduate, whether an applicant is offered a post-doctoral position or faculty position, whether a fellowship or grant application is funded, and whether a faculty member can continue in their position (often by earning tenure and promotion). Researchers who cannot satisfy these requirements can try applying elsewhere, or must leave academic research (aka "perish").
While this system may seem harsh, it's similar to how a donut shop owner must sell enough donuts to stay in business. In both situations, the public has limited resources to contribute, so there is strong selection for the most productive researchers and the most profitable donut shops. Competition can be fierce, and being successful typically requires long hours and a wide variety of skills. Some people thrive in this environment while others—by choice or necessity—will leave and apply their skills elsewhere.
How the internet changed the game
The internet has profoundly changed how scholarly work is disseminated. It used to be that individual researchers, research groups, departments, and universities had subscriptions to a limited number of print journals. Reading any part of a manuscript—including just the abstract—often involved searching the shelves of a departmental conference room or university library. Before copy machines became available in the 1960s, any detail you might need to remember had to be written down by hand. The process of submitting, reviewing, and revising manuscripts for publication was even more cumbersome.
All of these barriers meant that research results were essentially only available to a small number of privileged individuals at well-funded universities in developed countries. In the 1990s, the internet became generally available and changed everything. Anyone with internet access could search freely available websites like PubMed to access abstracts for published research articles. Reputable researchers without access to full-text articles could then contact authors and request a reprint copy by mail. Email and pdf documents soon made this process even easier.
Online databases were developed so that rapidly-expanding collections of data—such as DNA sequence information from the Human Genome Project—were accessible and searchable. The internet also made it possible for collaborators at different locations to easily share information and ideas.
This improved access allowed researchers at small colleges and at universities around the world to be active participants in a much wider array of research projects. Many of these institutions—and some of the local and national governments that supported them—saw research as a way to bolster their reputations and improve student education. As a result, many of these institutions adopted or strengthened "publish or perish" requirements for faculty members and even graduate students.
Journal proliferation
The ever-expanding body of knowledge has led to more specialized journals. The internet has made it far easier to access full-text research articles and to submit articles for review. Publishers have responded by introducing new online-only journals and open-access journals that are freely available to read, but that charge authors a fee to cover costs associated with publication. Many traditional journals are now online only, and some offer options for open access. Without the need to collect a certain number of articles before printing, many journals now post articles online shortly after they have been accepted. The pace of publishing has accelerated.
More researchers at more institutions are participating in research, and expected to publish. How often a researcher publishes has always depended on their abilities, the research field, and the resources available to them. For example, someone studying an endangered plant that only blooms once a year will have less data than someone who studies the fast-growing laboratory plant Arabidopsis. Someone working at a large, well-funded university with numerous research assistants should be able to publish more quickly than someone working on a similar project at a small institution with fewer resources. These differences get lost as so many researchers are rushing to publish, for fear of perishing.
Publishers have been happy to accommodate this rush by introducing even more journals. According to a 2019 study in GigaScience, there were ~12,500 active journals in the year 2000, and over 20,000 active journals in the year 2015. The average number of papers per journal has also increased, from 75 in 2000 to nearly 100 in 2015. The papers have gotten shorter, with an average length of 10 pages from 1980 to 2000, and an average length of 8.5 pages in 2014. The average number of co-authors has also increased.
Decreasing quality and the impact arms race
This huge increase in the number of papers being published is completely out of proportion with the actual increase in quality research being conducted. In order to compete for limited resources, scholars feel compelled to pursue lower-risk projects, publish inconclusive results, break a solid research project into multiple publications, or add their name to a project where they made just a small contribution. Others may use the same results for multiple publications, or fabricate data entirely. With so many journals—including predatory journals that charge a substantial publication fee but offer very little in the way of peer review—it's usually possible to get a weak paper published somewhere.
One measure of a paper's impact is its citation rate. According to a 2014 article in the Journal of Research in Medical Science, only 45% of the articles published in the top quartile of scientific journals are cited within 5 years of publication, including self-citation by the original authors. Papers published in lower tier journals would be expected to have even lower citation rates. Thus, it can be argued that many published papers primarily impact the authors' CVs, and do little to advance our collective understanding of the world.
Since the number of publications is no longer a reliable measure of scientific influence, various platforms have developed ways to calculate the "impact" of individual papers, authors, journals, and institutions. For example, an author's H-index is based on how often their most cited papers have been cited in other publications. A journal's impact factor reflects the average number of citations for articles published in that journal in the last two years.
Alternatives to "publish or perish"
When a scholar's H-index or other calculated metrics are used to make decisions about hiring, funding, or retention, the term "be cited or suffer" clearly applies. It's a natural extension of "publish or perish", and both terms deal primarily with how researchers interact with other researchers. Excellent researchers are expected to publish high-quality research in peer-reviewed journals in order to expand knowledge in their field, advance their careers, and bolster their institution's reputation.
Most institutions also value scholars who can effectively engage with the public. The terms "be discoverable or die" and "be visible or vanish" extend into this realm, and can refer to a scholar's visibility in a web search for their research topic. This visibility increases when scholars are active in platforms that are geared towards the general public. Traditionally, this might include giving lectures that were open to the general public, writing articles for newspapers or magazines, writing books that were accessible to the journal public, being interviewed on local radio or TV stations, or contributing to televised documentaries. Like traditional academic publishing, all of these options have gatekeepers who decide who is aired, published, or invited to speak on campus. Only a small number of scholars engage in these activities, but those who do might find that their scholarly voices are amplified, leading to more citations, more collaborations, and more scientific success.
With the internet, smartphones, and social media, anyone can now promote their own research and engage with the public. This can include putting together an informative website for your research group, Tweeting about exciting developments in your field, or posting explanatory YouTube videos or blog posts. While these activities are not traditionally considered scholarly work, perhaps they should be. If experts do not provide interesting and helpful information about relevant research topics on social media, non-experts will fill that space.
