While developing his theory of natural selection, Charles Darwin was horrified by a group of wasps that lay their eggs within the bodies of caterpillars, with the larvae eating their hosts alive from the inside-out.1
Darwin didn’t judge the wasps. Instead, he was troubled by what they revealed about evolution. They showed natural selection to be an amoral process. Any behavior that enhances fitness, nice or nasty, would spread.
Selection isn’t limited to DNA. All systems of inheritance, variation and competition inexorably lead to selection. This includes culture, and I’m one of a team of researchers at Arizona State University’s Institute of Human Origins who use a cultural evolutionary approach to understand human bodies, behavior and society.
Culture shapes everything people do, not least scientific practice – how scientists decide what questions to ask and how to answer them. Good scientific practices lead to public benefits, while poor scientific practices waste time and money.
Scientists vary. They might be meticulous measurement-takers or big-picture visionaries; cautious conservatives or iconoclastic radicals; soft-spoken introverts or ambitious status-seekers. These practices are passed on to the next generation through mentorship: All scientific careers start with years of one-on-one training, where an experienced scientist passes on their approach to their students.2 A successful scientist can train dozens of graduate students; meanwhile, poor strategies lead to an early career exit.
The Currency of Scientific Success
When scientists apply for jobs or funding, the primary way they compete is through their research papers: reports they write describing their work that are peer-reviewed and published in scholarly journals.
One of the sources of selection on scientists is how these papers are evaluated. Experts can provide detailed assessments, but many hiring or promotion committees use blunter metrics. These include the total number of papers a scientist publishes, how many times their papers are cited – that is, referred to in other work – and their “h-index”: a statistic that blends paper and citation counts into a single number. Journals are rated too, with “impact factors” and “journal ranks.”
All these metrics can incentivize some rather odd outcomes. For instance, citing your own past papers in each new one that you write can inflate your h-index.3 Some unscrupulous researchers have taken this to the next level, forming “citation cartels” where the members agree to cite one another’s work as much as possible, no matter the quality or relevance.
Recently there have been moves away from these simple-yet-flawed metrics. But without better alternatives, institutions simply put more emphasis on the raw number of publications, selecting for scientists to publish as much as they can, as fast as they can. Perhaps you’ve heard of the slogan “publish or perish,” or maybe even played the board game.
The Publishing Landscape
Scientists aren’t the only organisms in the scientific ecosystem. There are also publishers, the owners of the journals. Publishers live in an often-uneasy symbiosis with scientists, publishing their work, but also needing to make money off the process.
The traditional model was for journals to charge readers – or, more often, university libraries – subscription fees. This setup selects for journals to carefully vet their contents, as otherwise they will lose readers. Indeed, prominent journals reject the vast majority of submissions they receive.
The downside is that subscription fees block access for readers who can’t afford them. If you’ve ever tried to read an academic paper but been presented with a paywall, this is why.
Open Access Adaptation
The Open Access movement aims to make journal articles free for everyone to read and has led to many journals removing reader paywalls. But journals still need money, so most Open Access journals have swapped subscription fees for publication fees, paid by scientists on a per-paper basis.
This model allows anyone to read papers for free, but, as I have argued, it has also changed the selection pressures on journals, leading to some perverse outcomes.4
There are two ways for journals to succeed in this new landscape. For prestigious journals, they can leverage their reputation to charge large publication fees, sometimes over US$10,000 per paper.
For low-prestige journals, no one would pay such large fees. They must instead focus on quantity over quality. Like scientists, they must “publish or perish,” and publishers are already adapting to this new pressure – publishing more papers, opening new journals, increasing acceptance rates and expediting peer review.5
These changes created a new niche for scientists too, who are coevolving with the journals. An underhanded minority are adapting to laxer journal policies by using artificial intelligence to accelerate their research pipeline. The resulting papers are very low quality and so risk the authors’ reputations.6 However, until they are exposed, this strategy boosts research output and so brings rewards.
Alternatives
Publication fees aren’t the only model out there.
Diamond Open Access journals don’t charge fees at all and instead rely on donations.
Some scientists share what are called preprints, skipping peer review and putting their papers online for everyone to read for free. They may also publish them later in a conventional journal.
Academic society journals, which date back to the 17th century, often tie free publication to society membership and rely on interpersonal relationships and reputations to incentivize high-quality work.7
PCI’s or “peer community in’s” are groups of volunteer scientists aiming to wrest peer review away from journals entirely.
All of these are interesting options, and all would change the selective forces acting on both scientists and publishers. It makes sense to think about the evolutionary changes they could produce on the scientific landscape.
Why Scientific Evolution Matters
Darwin’s parasitic wasps reveal two truths: Selection is both unavoidable and amoral.
Whatever the domain, selection can lead to outcomes you might not like. For science, these might include the emergence of paper mills, mass retractions, citation cartels, fraud, excessive fees or bizarre AI-written papers.
But science can also do tremendous good: It produced modern medicine, discovered electricity and computing, and put people on the Moon. Like Darwin with his wasps, those of us who care about the scientific enterprise don’t need to limit ourselves to asking why some people do bad things. Instead, we need to ask why bad acts are selected in the first place and design better systems.
Don’t blame the player, redesign the game. If we can put better rules in place, evolution will do the rest.
Thomas Morgan, Associate Professor of Evolutionary Anthropology, Institute of Human Origins, Arizona State University
This article is republished from The Conversation under a Creative Commons license. Read the original article.
