The Doubter Is a True Man of Science

In 1847 a young Hungarian doctor named Ignaz Semmelweis noticed that women who gave birth in the Vienna hospital ward staffed by medical students died of fever far more often than those in the ward staffed by midwives. He doubted the common explanations of bad air and moral shame, and he observed that the students came to the maternity ward directly from the dissecting room. When he ordered them to wash their hands in chlorinated lime, deaths fell sharply. His colleagues were offended and dismissed him. The story shows what the topic claims, that the doubter, and not the believer, is the true man of science.

The claim rests on the character of scientific method. A belief that cannot be questioned is a dogma, whatever its subject, and science is distinguished from other ways of knowing by its readiness to test its beliefs. Karl Popper argued in the 1930s that a theory is scientific only if it makes claims that could, in principle, be shown false. The scientist is a person who sets up the experiment that might embarrass his own hypothesis and who is pleased, in a sober way, when it survives.

Descartes made doubt the foundation of modern philosophy. In the Meditations of 1641 he resolved to reject everything that could be doubted, to see whether anything remained that could not. What remained was the fact of his own thinking, and from it he tried to rebuild knowledge. His method has been criticised, but it set a pattern for later thinkers, which was to begin with a challenge to what is taken for granted and to accept only what survives it. The Indian tradition of vada, formal debate, in which the Nyaya school developed rules for testing claims, shows the same habit of mind.

The history of physics offers many instances. By the late nineteenth century most physicists believed that light travelled through an invisible medium called the ether. In 1887 Albert Michelson and Edward Morley tried to detect the earth's motion through it and found nothing. Their failure was puzzling, and it was the puzzle that led Einstein to the theory of relativity in 1905. Then in 1919 Arthur Eddington's expeditions measured the bending of starlight during an eclipse and found the result that Einstein's theory predicted and Newton's did not. Doubt had been tested by observation, and the old certainty gave way.

Doubt had been tested by observation, and the old certainty gave way.

The courage that doubt requires is sometimes physical. Galileo, who doubted that the heavens were as Aristotle and the Church described them, was tried by the Inquisition in 1633 and spent his last years under house arrest. Doubt is uncomfortable for communities as well as for individuals, because it threatens the order built on settled belief. Societies that protect the doubter, with academic freedom and a free press, tend to produce science, and those that punish him tend to lose it.

Yet not every doubter deserves the title. In the 1950s tobacco companies recognised that evidence linking cigarettes to cancer was growing, and they funded research and public relations to suggest that the matter was still open. The historians Naomi Oreskes and Erik Conway showed in Merchants of Doubt that the same techniques were later used to delay action on acid rain, the ozone layer and climate change. Their aim was not to find the truth but to delay a decision, and their doubt was never allowed to resolve into a conclusion. This is the counterfeit of the scientist's doubt, which seeks an answer and not a delay.

A genuine doubter differs in several ways. He doubts according to evidence, in proportion to how well a claim has been tested. He gives reasons that others can check and accepts the result if the test goes against him. And he applies his doubt to himself. Einstein, the greatest doubter of his century, doubted quantum mechanics until his death, famously objecting that God does not play dice, and the experiments of later decades went against him. A true scientist would regard this as part of the process, since the standards he set for others now applied to him.

Thomas Kuhn observed in 1962 that most scientists, most of the time, do not doubt the framework in which they work, and that revolutions come only when anomalies pile up beyond what it can explain. The true doubter is therefore a rarer figure than the topic suggests, and ordinary science runs on a settled consensus that few are willing to disturb.

Doubt also has costs, and a person who doubts everything cannot act. Doctors must treat patients before the final evidence is in. Governments must decide on vaccines, floods and wars with incomplete information. The ancient sceptics of the Greek schools recommended suspending judgement on all matters, and found that this brought peace of mind, but peace of mind is not the aim of the laboratory. Science needs provisional belief as much as doubt, a willingness to build on what has survived so far while remaining ready to rebuild.

Institutions help to discipline doubt. Peer review, replication and the open publication of data allow a community to check what individuals might overlook or want to overlook. Even so, the system is imperfect, and the replication crisis that emerged in psychology and medicine in the 2010s showed that many published findings could not be reproduced. The discovery came from scientists doubting their own field and doing the work to check it, which supports the topic's claim more than it undermines it. A discipline that can doubt itself in public has a chance of correcting itself.

For the public, the lesson is a difficult one. A citizen is not able to repeat the experiments, and must decide whom to trust. The test is whether a source invites scrutiny, publishes its methods and changes its mind, and not whether it sounds confident. A confident scientist deserves less trust than one who explains what would change his view. This leaves a question that science cannot answer for itself, about who is to doubt the doubters when they all agree, and whose agreement is itself a possible error that only a further doubter could discover.