Nick Spencer discerns the truth in scientific lies 27/06/2022

On 5 August 2014, the Japanese biologist Yoshiki Sasai hanged himself in his offices at the Riken Institute’s Centre for Developmental Biology, in Kobe. He had been supervising a young researcher who had made an astonishing breakthrough in stem cell research, which had generated much attention, including two papers for the prestigious journal Nature. Frustratingly, however, no other laboratory around the world was able to replicate her results.  

People became suspicious. The Riken Institute commissioned an investigation, and the findings were damning. The researcher’s work was sloppy, her lab notes were opaque, her data were manipulated, her work lacked integrity and humility, and her supervision was inadequate. The researcher was compelled to retract the article. Nature was embarrassed. Professor Sasai was mortified. 

There are two ways to read this story. The first is that it is an example of science doing what it does well. No scientist is perfect, intellectually or morally. Science is the slow process of winnowing, checking and rechecking people’s work, and outing that which fails to pass muster. 

The second is that it is an example of science failing to do what it claims to. Of course, no scientist is perfect. But you don’t need to be perfect to take good notes, show integrity, and avoid manipulating data. I guess it all depends on whether your test tube is half full or half empty.  

* 

Whichever way you read the sad story of Prof Sasai and his corrupt researcher, the real problem is that it is not a one off. In his perceptive and well–balanced book Fraud in the Lab, the investigative journalist Nicolas Chevassus–au–Louis, charts multiple stories of, well, fraud in the lab. 

There is the well–known tale of Woo–Suk Hwang, the South Korean biologist whose apparent success with cloning was shown to be illusory, not only ethically questionable, but straightforwardly corrupt – he had retouched photographs and faked results – and un–replicable. Hwang was fired, and sentenced to two years in prison for fraud, embezzlement, and violating bioethics laws. 

And then there is the tale of the Dutch psychologist Diederik Stapel, who was forced to retract 55 of his 130 journal articles because it was revealed he had basically made up the data himself. And the tale of American Neuroscientist Marc Hausser, who was forced to resign from Harvard, despite having several times been elected “most popular professor of the year”, because the Office of Research Integrity discovered repeated examples of data fabrication in his work. And German physicist Jan Hendrik Schon, who had published multiple widely celebrated articles in both Nature and Science and was tipped for a Nobel Prize, until it was revealed that he had used mathematical functions to generate plausible results for experiments he had never conducted. And so on, and so on. 

In one respect, none of this is new. Chevassus–au–Louis begins his book by talking about the mathematician, Charles Babbage, whose book Reflections on the Decline of Science in England fingered hoaxing, data forging, and trimming, was published in 1830(!) Nor is it evenly spread. In spite of Schon’s work, physics is largely spared such fraudulent activity, as is mathematics. Biology, chemistry, psychology, the social sciences, and especially medicine have a much more serious problem. 

However, it is growing, with more and more examples of corrupt science emerging. According to Chevassus–au–Louis, “the proportion of retracted articles [in biology] is ballooning, increasing elevenfold between 2001 and 2010”. (10) Other disciplines show comparable increases.  

It is very important to get two important caveats in place here. Firstly, we are starting from a low base. To give some concrete figures: the rate of retractions in the journal Cell is only 0.16%, Science 0.08%, Nature 0.05%. Even were these numbers to increase elevenfold again, they would still be tiny. Only in fantasy land does this constitute an epidemic. 

Second, not all retractions are fraudulent. Indeed, by the calculations of one study, only about a quarter are. Even if we are to think, as Chevassus–au–Louis does, that this may only be the tip of the iceberg – he reports findings that say that 14 per cent of scientists say they are aware of colleagues who commit fraud – it is still several light years away from being ubiquitous. 

These are important caveats because there are many fringe groups who want to discredit science at the moment, for ideological (or frankly nutty) reasons. Anti–vaxxers and climate change deniers do not need anyone to give them ammunition. 

But that is precisely the irony in all this. Pointing out the fallibility of science can only be considered “ammunition” in a culture in which science is assumed to be infallible. And modern science, from its very origins in the work of Francis Bacon in the early seventeenth century, is predicated on the idea of human fallibility. As humans, beset by sin, we are unable simply to think our way to the truth. Instead, we must feel and experience our way there. Experiment, in the seventeenth century, was commonly used in association with prayer, experimental prayer pertaining to personal, experiential, as opposed to formalised and ritualised prayer. 

Science only finds itself in this weird situation, in which its opponents think they have scored a hit against it when they point out its errors and flaws, because it has been culturally elevated to the position of unimpeachable authority and certainty. But, irrespective of whether you think it merits that position, it is only there because it is founded on uncertainty, essentially a process of organised scepticism. It is trustworthy because it doubts. 

For many people, this crystallises the difference between religion and science. Thus, some of our interviews in the Theos/ Faraday Science and Religion project told us, religion is a belief system, science is a doubt system.  

It’s a good aphorism but doesn’t bear much weight. After all, belief without (at least the capacity for) doubt is no longer belief. It’s false certainty. For all that religion is about faith, the idea that it is simply about blind, mindless, don’t–care–about–the–evidence faith is a fiction of polemicists.  

Conversely, doubt must stop somewhere otherwise the process simply eats itself, ending up with the orchestrated self–harm of denialism. Doubt systems usually end up believing something. 

The process of discernment and wisdom should lie at the heart of religious faith. As one of my colleagues put it to me, organised scepticism is to science what discernment is to religion. Or as St Paul once wrote to his over–enthusiastic audience in Thessalonica, “test everything [but] hold on to what is good.”  

Nick Spencer is Senior Fellow at Theos 

Fraud in the Lab: The High Stakes of Scientific Research by Nicolas Chevassus–au–Louis is published by Harvard University Press.