UNTIL the invention of the printing press in the 16th century, scientists could communicate only with friends and colleagues. After this technological revolution they were able to publish books and journals, with profound impact on the development of science. The pioneering Journal des Sçavans (Journal of the Learned), appeared in France in 1665. A decade later, this journal published the calculation of the speed of light by Ole Romer. The fastest thing in nature was communicated at a speed previously unavailable to scientists.
Over the next few hundred years, scientific journals became increasingly important, overtaking books as the primary means of scientific communication. Without journals, science would not have developed as it did, and those early journal editors and printers are unsung heroes of scientific progress.
However in the mid-20th century, academic publishing took a turn for the worse. Starting with Robert Maxwell and his Pergamon Press, commercial publishers understood that the monopoly in scientific publishing could be very profitable. When a paper is published in only one journal, major university libraries must subscribe to that journal, no matter how expensive it is, to ensure that their scientists can access the whole scientific literature.
A 1992 survey of journals in the field of statistics showed that most society journals charged libraries less than $2 per scientific research article, while the most expensive commercial journal charged $44 per article. At the time, that was more for a single journal article than the average price for an academic book.
It has worsened since. Being both the producers and consumers of scientific articles, universities pay an enormous amount for journals which contain articles that are both written and peer-reviewed by their own scientists, which they provide to journals free of charge. As a result, scientific journal publishers have huge profit margins reaching almost 40 per cent.
The academic publishing problem is not only about cost and access. Throughout most of history, it was the importance and quality of the scientific article that mattered, not the journal in which it was published. Scientists didn’t care much about journal prestige, but they wanted to reach as many fellow scientists as possible, which was best accomplished through journals with many subscribers. This created a hierarchy among journals. A large flow of submissions to widely circulated journals led to high rejection rates, which in turn made them more prestigious to publish in.
When hiring and promoting scientists, it can be tedious and time-consuming to read and evaluate all the papers of all the different candidates. To save time, the prestige of the journal in which the authors have published is sometimes used as a surrogate for article quality. This may seem strange to non-scientists, but depending on the field, every young scientist knows that the acceptance or rejection of a research article by Science, the Lancet, Econometrica or Annals of Mathematics can make or break a career. This ‘incentivises careerism over creativity’.
As eloquently expressed by former National Institutes of Health director Harold Varmus and colleagues: ‘The inflated value given to publishing in a small number of so-called “high impact” journals has put pressure on authors to rush into print, cut corners, exaggerate their findings, and overstate the significance of their work. Such publication practices . . . are changing the atmosphere in many laboratories in disturbing ways. The recent worrisome reports of substantial numbers of research publications whose results cannot be replicated are likely symptoms of today’s highly pressured environment for research. If through sloppiness, error or exaggeration, the scientific community loses the public’s trust in the integrity of its work, it cannot expect to maintain public support for science.’
Peer review has a long and rich history, and it is an indispensable part of scientific discourse, as evidenced by many scientific controversies and discussions. Scientific peer review takes many forms, including published commentaries, positive or negative citations, and discussions at scientific meetings. In the 20th century, journals initiated a system of anonymous, unpublished peer reviews. It was costly to print and ship paper journals, so not everything could be published, and editors started to use anonymous reviewers to help determine what to accept or reject.
This led to the strange idea among some scientists, where ‘peer-reviewed research’ became synonymous with research published in a journal that uses an anonymous peer-review system to determine what science should be published, ignoring the many traditional forms of open and non-anonymous peer review.
Universities and other research institutes, as well as research funders, have an intrinsic need to evaluate the science and scientists they employ and support. By relying on journal prestige instead of article quality, they have outsourced parts of their evaluation to unknown people without seeing the actual reviews. Such a system is ripe for mistakes and misuse.
Compared with good research, questionable manuscripts require the effort and time commitment of more reviewers, as they are more likely to get rejected and resubmitted. Even fatally flawed manuscripts are typically accepted by some journal eventually. This gives the research a stamp of approval of being published in a ‘peer-reviewed journal’ but without readers having access to those earlier critical reviews. Would it be better if those flawed research papers were published by the first journal together with the critical reviews, so that readers could have learned about the problems with the studies?
While we cannot prevent bad science from being published, what is needed is open, robust, and lively scientific discourse. That’s the only way to seek scientific truth.
A longer version of this article appeared in Brownstone Institute on October 14, 2025, and is republished by kind permission.
