BOOK REVIEW: Why Trust Science?

By Naomi Oreskes. Princeton, NJ, USA: Princeton Univ. Press, 2019, 360 pp.

Reviewed by Jacob Ossar

Anyone paying even the slightest attention to the news in the last few years would have been hard-pressed to avoid reports of wildfires, floods, hurricanes, heat waves, and other natural disasters. The severity and frequency of these extreme weather events raise urgent questions about the degree to which they are related to climate changes and what measures we should take in response. Meanwhile, the COVID-19 pandemic has forced people worldwide to grapple with questions about the efficacy of measures like masks, social distancing, and vaccines in controlling the spread of the virus or mitigating its effects. Most of us look to science for answers to these questions, but not everyone accepts the answers scientists give. Some dispute particular scientific claims, but some go farther and challenge the trustworthiness of science as an enterprise. Naomi Oreskes, a professor of the history of science and affiliate professor of earth and planetary sciences at Harvard University, addresses these doubts and challenges in Why Trust Science?

Oreskes’ answer to the book’s eponymous question is that science is trustworthy to the extent that the social process by which scientists vet research findings and reach (or fail to reach) a consensus about them is open to a diverse community of scientists with ample opportunity to make objections and critiques and have that feedback taken seriously. She devotes slightly under two-thirds of the book to laying out her argument. In the remainder, she models what she has argued are best practices for seeking knowledge by including critiques of her argument from historian of science Susan Lindee, philosopher of science Marc Lange, climate experts Ottmar Edenhofer and Martin Kowarsch, and political scientist Jon Krosnick, followed by her replies to their various objections.

To set the stage for her conclusions, Oreskes gives us nothing less than a short course in the philosophy of science. While her summary is necessarily brief and incomplete, she cogently highlights key threads in the development of (largely) Anglo-American philosophy of science from the 19th century to the present. Although the goal of this potted history is to show where and how other influential answers to her question have fallen short, her engagement with the thinkers who serve as her intellectual foils is serious and sympathetic.

Oreskes begins her brief history in the 18th and early 19th century, with the establishment of scientific honor societies like the Royal Society, the Académie des Sciences, and, later, the U.S. National Academy of Sciences. These bodies served to identify individual “men of science” whose findings could be treated as authoritative. However, as any first-year logic student could tell you, the mere fact that a Great Man claims something is true is not a good reason to accept it. To the extent that it is rational to take a Great Man’s word for something, it is because we believe that he has earned his authority by having some reliable means of discerning or discovering what is true.

Oreskes gives us nothing less than a short course in the philosophy of science.

Many of us were taught in grade school that the scientific method works something like this: start with a question, make observations, formulate a hypothesis, conduct experiments to test the hypothesis, analyze the results, and then devise a theory that explains them. While there is no doubt that this is a scientific method, Oreskes denies that it is the scientific method: observation does not confirm the hypothesis that all scientists use this method and only this method. She endorses Paul Feyerabend’s view, expressed in his book Against Method, that if you look at what actual scientists in different fields do, you will see that they employ a wide variety of methods depending on what best suits the task at hand. Science is a social activity that seeks to arrive at a consensus about what is true of the natural world by weighing a variety of different factors. This includes things like quantity of evidence, experimental design, and sample size, but also, because scientists are human, reputation, individual affinities or rivalries, prejudices, and so on. One influential way of putting this point is to say that scientific knowledge is “socially constructed.”

You might think that if science is unavoidably social, then it is also unavoidably subjective and hence unable to get at the truth reliably. In a nice bit of philosophical judo, Oreskes argues that the socially constructed character of scientific knowledge actually provides resources for a strong defense of science. We know that individual scientists see things from their own point of view, which necessarily influences their judgment. But if the community of scientists includes enough members with varied points of view, by reviewing and vetting one another’s results, the scientific community can reach a consensus that corrects for blind spots any given scientist may have. The social nature of science, with its strong emphasis on peer review and the replicability of results, allows scientific communities to approach objectivity much more closely than individuals working in isolation could possibly hope to do, no matter how solid the method they employed.

Science is a social activity that seeks to arrive at a consensus about what is true of the natural world by weighing a variety of different factors.

Oreskes argues for an ideal of science produced by diverse communities of inquiry where different viewpoints receive a fair hearing, yielding a strong, trustworthy consensus. But, as she would be the first to acknowledge, actual existing scientific communities often fall short of this ideal in various ways. How much should this influence our trust in them? According to Oreskes, our trust in science is not a binary matter, but instead a probabilistic one. A scientific consensus is trustworthy to the extent that it comes from a scientific community that approaches her ideal; the more a community deviates from it through failures of diversity, and so on, the less trustworthy its findings are.

We know of historical cases where the consensus of a particular scientific community has been overturned by later developments. We also know that no scientific community is perfect. So, how can we rely on current scientific knowledge when we cannot be sure it will stand the test of time? Those skeptical of the scientific consensus on climate changes or some other issue might argue that since scientists have been wrong before, why should we trust them this time? Oreskes devotes the second chapter of her book, “Science Awry,” to case studies of apparent scientific consensus going wrong, arguing that in each case, these failures raised red flags that were evident not just in hindsight, but also at the time. Being sensitive to factors that should lead us to view findings skeptically, withhold judgement pending further research, and so on will also help us to feel confidence in cases where these red flags are absent.

One troubling case Oreskes discusses in this chapter is the embrace of eugenics in the early 20th century by scientists (as well as by politicians and other non-scientists). Oreskes argues that this is not merely a case of the misuse of science for political ends; there were scientists who presented eugenics as a straightforward consequence of Darwinian evolutionary theory. Skeptics of science claim that this view represented the scientific consensus at the time, but Oreskes—citing strong contemporary criticisms of eugenics by sociologist Franz Boas, anthropologist Margaret Mead, geneticists J. B. S. Haldane, J. D. Bernal, Julian Huxley, and Herman Muller, and others—shows that this is far from having been the case.

The social nature of science, with its strong emphasis on peer review and the replicability of results, allows scientific communities to approach objectivity much more closely than individuals working in isolation could possibly hope to do.

We might be tempted here to distinguish between “purely scientific” objections to eugenics and political or value-based ones. But, if we take Oreskes’ account seriously, it is not clear how tenable this distinction is. As we have seen, Oreskes rejects the idea of any unitary scientific method and instead places her trust in a diverse scientific community cancelling out distortions that might arise from individual perspectives through peer review. So, a scientist’s gender or political orientation might wind up being just as important scientifically as her attention to detail, sense of curiosity, mathematical sophistication, and so on. Thus, for example, Herman Muller’s socialist convictions led him to be especially sensitive to the ways in which the purported evidence in favor of eugenics could not be imputed to purely genetic factors given prevailing social conditions.

Even if Oreskes is right that in cases of science gone awry, there are red flags for contemporary observers to notice, actually finding out about those red flags in real time might be difficult for non-experts. To be fair, Oreskes titled her book Why Trust Science? not How to Trust Science. Still, it is important for her defense of science that non-experts be able to identify bona fide experts with some degree of confidence. This is related to what early 20th-century philosophers of science called the “demarcation problem”: how to distinguish science from superstition, pseudo-science, metaphysical speculation, and the like. According to Oreskes, what makes scientists scientists is not any particular thing they do as individuals, but instead their membership in a scientific community. Non-scientists can tell who is a scientist by looking for signs of community membership like tenure at a research institution, prizes and awards, journal publications, and so forth. But all of these signs of expertise are bestowed by other scientists, and so the question of how we can tell who is a scientist and who is not recurs, suggesting a degree of circularity.

Actual existing scientific communities often fall short of this ideal in various ways.

For Oreskes, the way out of this circularity is to insist that “the social markers of expertise are evident to non-experts.” So, for example, we can look at a putative expert’s institutional affiliation and “it is (or should be) easy to distinguish a research institution—like Princeton University or the Lawrence Livermore National Laboratory—from policy-driven think tanks, such as the American Enterprise or Discovery Institutes. The fact that journalists often fail to make such distinctions has more to do with deadlines than with epistemology.” The final sentence suggests that things are perhaps not as easy for non-experts as she makes out, since many of us rely on journalistic accounts to find out what the experts are saying and such accounts may elide or omit crucial details even if the journalists responsible for them could and ought to be more careful. Furthermore, a distressing number of people have been taken in by wholly fraudulent online diploma mills, so the familiarity with the landscape of higher education necessary to determine whether an institution is a reputable college or university might not be as widespread as Oreskes hopes.

In any case, even sophisticated laypeople who can confidently distinguish between a scientist and a quack still face urgent policy decisions that hinge on scientific findings we may not fully trust. What are we to do when we are unsure whether the consensus of the scientific community is correct? For example, do we believe the scientific consensus about the reality of and mechanisms underlying climate change? Oreskes suggests that we should think of the problem as an instance of Pascal’s wager, asking ourselves how the risks of rejecting the science if it turns out to be true compare to the risks of accepting it if it turns out to be wrong.

Oreskes writes that “it would be absurd for me to expect that articulating the reasons for trust in science would alter the positions of climate change deniers. What is not absurd is the hope that for some readers, this book will answer legitimate questions, even if those questions have at times been raised by people with a political agenda with which I strongly disagree.” Such readers will find that Oreskes’ writing is admirably clear and easy to follow.