Why Trust Science?, by Naomi Oreskes (Princeton and Oxford: Princeton University Press, 2019, ISBN 9780691179001). 360 pp. Hardcover, $24.95.
Naomi Oreskes’s new book, Why Trust Science? asks a question that secular humanists should be able to answer. We are often touting the trustworthiness of science, or at least scientific reasoning, over faith. But exactly how trustworthy is science, and what is the basis for that trust?
Oreskes is professor of the history of science and affiliated professor of earth and planetary sciences at Harvard University. One of her previous books is Merchants of Doubt: How a Handful of Scientists Obscured the Truth on Issues from Tobacco Smoke to Global Warming. She cites Merchants of Doubt often in her current book, making me wish I had read that one first. I also wish her new book had been written after the COVID-19 pandemic, which seems certain to have an impact, for better or worse, on the public’s attitude toward science.
A core principle of science has been that all scientific knowledge, no matter how fully verified, is provisional. This is, of course, one of the things that distinguishes science from religion. Science offers provisional truth based on evidence, while religion offers absolute truth based on no evidence at all. In any case, the willingness of the scientific community to change its mind based on new facts would seem to be an admirable quality, one that would inspire trust. But it also presents a problem. As Oreskes puts it, “How are we to evaluate the truth claims of science when we know that these claims may in the future be overturned?” And, as we know, the truth claims of science are all too often wrong. Oreskes has some laughable, in fact riotous, examples to share.
At one time, science was trusted because it was the province of “great men.” One of these greats was Dr. Edward H. Clarke (1820–1877), who argued based on his “limited energy theory” that over-stressing one physiological system, such as the brain and nervous system, would divert energy from other systems—most specifically, according to Clarke, the female reproductive system. Simply put, the demands of higher education would cause a woman’s uterus and ovaries to shrink. Clarke did not specify which parts of the male anatomy were subject to shrinkage from rigorous intellectual activity.
This conclusion was supposedly based on a “study” by Clarke of seven college-educated women. In fact, it had more to do with the eugenics movement. Clarke feared that a declining birth rate among native-born women, along with the massive influx of immigrants, would lead to the race being “propagated from its inferior classes.” As a basis for confidence in science, trust in great men seems hopelessly misguided. We can’t judge 1873 by 2020 standards, but it is still dumbfounding that Dr. Clarke’s 1873 book, Sex in Education; or, a Fair Chance for Girls, sold some twelve thousand copies over the course of nineteen editions in the following three decades.
When we ask why we should place our trust in science, today we would probably answer that science includes a principle and process of verification. Scientific conclusions, by definition, are based on facts objectively verified. But again, new facts can lead to abandonment of what were once the accepted facts, and there is the further problem that empirically based verification is not always possible.
Apparently, a couple of years ago there was a huge flap over the value of flossing your teeth. I know flossing is not the sexiest thing you can think of doing, but it seems pretty obvious that floss can go where the bristles of your Oral-B dare not tread. In any case, it was argued that there was no hard evidence that flossing does any good. The problem here, of course, is that collecting hard evidence by way of randomized trials would create a serious ethical dilemma. If flossing does work, you would be putting your non-flossing subjects at risk for periodontal disease.
The dental floss crisis was not really a case of science gone wrong but of one reporter jumping to conclusions and science getting the blame. The evidence in favor of flossing was not perfect, but that did not mean there was no evidence at all. And, as Stephen Macedo writes in his introduction to Why Trust Science?, nonexpert opinion may have a role to play in science in a democracy.
However, Oreskes reviews other, more egregious failures, directly attributable to the scientific experts. Should we, for example, trust science when it took about four decades for the theory of continental drift to be accepted? The history of eugenics is surely complex, but it is not hard to see how such episodes could undermine the average person’s trust in science. Why were women’s reports of depressed mood after taking the birth control pill ignored for years? This was despite the fact that the relationship between reproductive hormones and mood-altering neurotransmitters was well understood—so well, in fact, that clinical trials of a male contraceptive injection had already been halted due to, yes, an increase in mood disorders.
Enter Karl Popper and his theory of falsifiability. For any hypothesis to be true, there has to be something that can count against it. This is a problem for the God hypothesis. There appears to be no actual or possible outcome that would throw doubt on God’s existence; therefore, the hypothesis is false. Regarding science itself, however, Popper recommended that the scientist’s job was not to prove theories to be true but to set out to refute them. But Popper may have pushed fallibilism a bit too far. If our theories, Oreskes asks, “are not only soon to be refuted, but should be refuted, then why believe any of it?”
There are other candidates for assuring that our confidence in science is properly placed. Replication would be one, but how many scientists want to build a career on replicating the work of others? Peer review is another, but peer review presumes that science can be trusted to police itself. Oreskes proposes that consensus among scientists is another source of trust, but unfortunately consensus has traditionally been a consensus of white males, with all the biases and prejudices inherent in any homogenous group. What is needed is a more diverse scientific community, one in which no single bias can predominate.
Oreskes’s book is intended to argue for greater diversity in science, but she also believes that what is needed is an open discussion “of the role of values in both science and policy.” Much of the distrust in science is based not on science but on economic interests and ideological commitments—that is to say, on values. It may be a mistake, therefore, for scientists to argue that in their quest for objectivity, they remain value-free. As Oreskes asks: Why trust someone with no values? A better strategy would be for scientists to admit that values inevitably inform their work and to appreciate the ways in which they share values, even with their antagonists.
Why Trust Science? is built on a fairly familiar format. The introduction, by Stephen Macedo, prepares the reader perfectly for what is to come. (The book is also one in a series edited by Macedo of The University Center for Human Values at Princeton University.) The beginning essay, by Naomi Oreskes, is naturally the core of the book and is followed by responses from a variety of theorists and experts. The book then ends with a reply and then a brief afterword by Oreskes. Fortunately, the respondents used their time and space to address Oreskes’s ideas as opposed to beating their own drums, making the back and forth integral rather than tangential to the book.
Beyond being an ideal introduction to the history and philosophy of science, Why Trust Science? is a thoroughly reasoned account of the evolution of trust within the scientific community and the lay public.