A journal editor asked me to provide some quick comments on scientific journals, looking back on history and towards the next fifty years. Here is my response:
About the future of your journal as well as journals generally, the next fifty years is too long to prognosticate about. This is the case even though during the one hundred years of the 20th century, there wasn’t a huge amount of change in the way journals were published. Moving online turned out to be a wasted opportunity for most journals, which continued to have their length limits, their subscription models, and their practice of publishing only (distorted) summaries of scientific findings, rarely or never the actual raw data or the enabling technology (code, in the case of our research) behind them, with corrections/retractions continuing to be extremely rare, despite a large proportion of papers having major errors, as we now know thanks to meta-science. Subscription price increases way above inflation continued, and most academics continued to donate their labor to even the publishers with exorbitant profits.
This disappointing history indicates that we might expect little change in the future — enormous technological change (the development of the internet and move of communication to it) can have very little effect on conservative institutions such as the scientific establishment. But technological progress is continuing at a rapid pace, so by fifty years from now, artificial intelligence, and the robots they will control, may have completely taken over the planet! And even if they don’t reach that point by then, life will be very different. Moreover, looking at a shorter time horizon of the past twenty years in scientific publishing, there have been some fairly dramatic changes, although many journals (possibly including yours, apologies for my ignorance of it recently) have continued in an unperturbed manner.
A very slow increase in pressure from individual scientists and from public and private funders have propelled increases in open access and open science, catalyzed to some extent by the replication crisis. Meta-science, which diagnoses problems based on empirical data and proposes solutions, is now a flourishing field, whereas outside of medical journals it barely existed fifteen years ago. Initiatives like DORA and the Leiden Manifesto have gained lots of signatories, representing a shift, albeit a glacially slow one, towards real journal quality indicators like the TOP guidelines, rather than exclusively legacy factors like that “everybody in the past submitted there so we’ll submit there, regardless of the quality of their processes and the transparency of the results published there.”
It’s a good bet that the slow shifts I’ve mentioned will continue for the foreseeable future, although likely they’ll continue to be too slow to avoid enormous continued waste in research spending and the undermining of action to address the climate emergency because a lack of reform at journals undermines the trustworthiness of scientific results in many areas, including climate research.
In summary, although progress will be too late for much of the ongoing damage from bad and irreproducible biomedical research and climate research, as well as hyped-up and false-positive psychology findings, I’m expecting that many journals, hopefully including yours, will adopt additional policies to encourage preregistration, Registered Reports, open data, corrections, computational reproducibility checks during peer review, open peer review, reform of authorship guidelines to recognize the increasing amount of teamwork involved in science, etc.
