Earlier this year, Van Pelt submitted a completely fabricated article that was accepted without question, even though he explicitly wrote that no experiment had ever taken place.

Researchers will do anything to expand their publication list, and these journals exploit that drive, with almost no oversight in place. Van Pelt calls the article mills—often based in Iran or China—“truly criminal.” For a fee, sometimes with AI assistance, they produce entire articles on behalf of scientists who don’t have the time to write themselves. This is particularly common in non-Western countries where internal university oversight is weak or nonexistent. Most of what appears in those articles, Van Pelt says, is complete nonsense.

Why is that such a problem? Some might say: let researchers sort it out among themselves.

“Because fellow researchers don’t always realize the data are fake. Or that an experiment has been inflated—for example, that there were only ten test subjects instead of the hundred or thousand the article claims. But those results are still included in meta-analyses, which in medicine are used to form clinical guidelines. I spoke several times with Dutch professor of gynecology Ben Mol at Monash University in Australia, who is outraged about this and trying to stop it. He says: ‘Women and babies are dying because we allow this to happen.’ Faulty data are a kind of fake news that is polluting the entire scientific literature.”

How can science become healthy again?

“First of all: my book focuses on the excesses. A lot is going well, and there’s a large gray area between someone like Diederik Stapel, who fabricated entire studies, and someone who follows every rule to the letter. But I think we need to partly return to the old funding system, with less money funneled through NWO and more distributed directly to universities. We need to end the practice of grants that only last four years—the length of a PhD project. That way, universities could employ more people permanently.

Currently, 56 percent of researchers are on temporary contracts, and for PhD candidates it’s nearly 100 percent, even though they do most of the actual research work. If universities had more permanent jobs to offer, competition would ease, workload would decrease, and the perverse incentives for fraud would disappear. At the same time, we need to limit the influx of PhD candidates. There simply aren’t enough positions for them, and society has no real need for such large numbers of narrowly trained specialists.”

In your book you also propose some form of inspection service for science.

“Yes, I argue for better oversight. Many professions we value highly have inspectorates. Think of the Netherlands Food and Consumer Product Safety Authority or the Health and Youth Care Inspectorate. Why doesn’t science have something similar? Someone should occasionally check how data are structured and what analyses have been carried out.

If you insist on using the sports analogy: why do cyclists use doping? Because they want to be the best, and competition pushes them over the edge. Scientists also want to score and push boundaries. In cycling, that’s why doping checks were introduced: every day, someone can show up at your door asking for a urine sample. It doesn’t have to be that extreme in science, but it would help if researchers thought more often, ‘I’d better make sure everything is in order.’”

But universities already have scientific integrity committees, and there’s also a national body, LOWI.

“Those committees only investigate once someone has sounded the alarm. Until then, everyone assumes everything is fine. That’s exactly why Diederik Stapel was able to continue for so long. Moreover, the judgments these committees issue are only advisory. An executive board can decide what to do with them. I’d say: make such rulings binding. A researcher who crosses the line could, for example, be barred from research for two years, or in more serious cases, dismissed. Right now, that almost never happens.

But oversight can also be more low-threshold: implement a four-eyes principle. Before a paper is submitted, have someone else check whether all calculations and analyses are correct. Mistakes creep into research all the time, often because someone doesn’t fully understand the statistics.”

Restoring trust in science

Van Pelt has also thought about solutions for the problem of fake publications. Serious scientific journals should take swifter action when fraud detectives, like the aforementioned Ben Mol, raise concerns about a paper. In addition, academia should reclaim publishing itself. In the past, professional associations had their own journals, but with the rise of commercial publishers like Springer Nature and Elsevier, most of those disappeared. “Here in Nijmegen, Radboud University Press could take on that role. In some fields, like linguistics, researchers are already starting their own journals again and quitting editorial work for commercial publishers. Those are promising initiatives.”

In Van Pelt’s view, much is at stake. Public trust in science in the Netherlands remains high. A recent report from the Rathenau Institute found that people give science a score of 7.4 out of 10, higher than all other institutions studied, including the judiciary and the House of Representatives. “Science still has a lot of credit, and rightly so. But to maintain that trust, it needs to take its darker sides more seriously.”

This article was translated using AI-assisted tools and reviewed by an editor.