Back to Articles

I Am Certain…Just Not One Hundred Percent

By Lauren Kantor Gorman MD
September 24, 2018 | Comments

Editor’s Note: Dr. Lauren Kantor Gorman is a New York City psychiatrist, mother of Critica co-founder Sara Gorman and wife of Critica co-founder Jack Gorman.

What happens when scientists can’t say they are completely sure?

On August 1, 2018 the New York Times Magazine took the unusual step of devoting an entire issue to one issue: climate change. “Losing Earth: the decade we almost stopped climate change. A tragedy in two acts,” by Nathaniel Rich, with photographs by George Steinmetz, has been hailed as an incredibly well-researched, nuanced, and gripping account of what we could have done—and failed to do—to prevent runaway global warming in the decade between 1979 and 1989.

        Rich primarily blames human nature for the lost opportunity. Had we heeded warnings that human-driven carbon dioxide and methane emissions were driving up the earth’s temperature back in 1979, we might have averted what is now the looming climate catastrophe. Unfortunately, he laments, we humans have a hard time grasping risks that are in the distant future. We naturally default to ignoring them in favor of more immediate worries. Here Rich is on firm scientific ground, for scientists have indeed demonstrated this “immediacy bias” in the way we assess risk and make decisions. Jack and Sara Gorman have written about this kind of evolutionarily conserved cognitive default in assessing risk in their book Denying to the Grave (Oxford, 2016).

        The New York Times piece also garnered considerable criticism, mostly from climate activists who believe Rich wrongly placed the main blame for climate change denial in the early 1980s on the general public. For these critics, he should have placed much more blame on two villains, the fossil fuel industry and the Republican Party.

        Clearly, powerful forces were at work to suppress concern and action about climate change 30 years ago, including human nature, corporate greed, and political ill-will. At Critica, we of course wonder why scientists who knew then that carbon dioxide emissions from burning fossil fuels were posing an imminent risk of climate meltdown were not able to convince politicians, CEO’s, and the rest of us that we were—and even more so are now—in terrible danger of destroying human civilization. The scientific information was clearly at hand then as it is now, as Rich nicely explains in his magnum opus, and the data irrefutable in climate scientists’ minds. “All the facts were known, and nothing stood in our way. Nothing, that is, except ourselves,” he writes. Why weren’t scientists able to persuade us that immediate action was needed?

Stuck on a Paragraph

        In reading through Rich’s article, which despite the many reservations I believe deserves a Pulitzer Prize, I came across a fascinating exchange on page 26 among a group of climate scientists, activists and policymakers assembled at a meeting in October 1980. The group was charged with developing a plan for climate legislation to be presented to the recently created National Commission on Air Quality.

        At one point in the three-day meeting, Rich relates, the group became “stuck on a sentence in their prefatory paragraph declaring that climate changes were ‘likely to occur.’” We will reprint the dialogue that ensued exactly from the Rich article (go to page 26 to get the full names of the people quoted and their affiliations):

“Will occur,” proposed Laurmann, the Stanford engineer.

“What about the words: highly likely to occur?” Scoville asked

“Almost sure,” said David Rose, the nuclear engineer from M.I.T.

“Almost surely,” another said.

“Changes of an undetermined – “

“Changes as yet of a little-understood nature?”

“High or extremely likely to occur,” Pomerance said.

“Almost surely to occur?”

No,” Pomerance said.

“I would like to make one statement,” said Annemarie Crocetti, a public-health scholar who sat on the National Commission on Air Quality and had barely spoken all week. “I have noticed that very often when we as scientists are cautious in our statements, everybody else misses the point, because they don’t understand our qualifications.”

“As a non-scientist,” said McPherson, the congressman, “I really concur.”

        The two dozen experts in that conference room never got past that first paragraph and never made any recommendations for actual laws to rein in climate change. Instead, they got stuck on something that scientists struggle with all the time when trying to break out of their laboratories into the public square: how do we convey certainty when we aren’t completely certain?

Rejecting the Null Hypothesis

        The problem rests with the very basis upon which experiments are conducted in accordance with the scientific method. Scientists are trained to understand that what they are doing, no matter what field they are in, is trying to “disconfirm the null hypothesis.” Instead of saying “I am trying to prove that this is true,” scientists always start with the premise “I am trying to prove that this is not true.” According to the scientific method, experiments begin with the creation of a hypothesis. In the case of climate change it might be “I hypothesize that there has been an increase in carbon dioxide emissions since the introduction of the gas-burning engine.” But the way the experiment actually works is to take the reverse of that hypothesis— “there has been no change in carbon dioxide emissions since the introduction of the gas-burning engine”—and try to prove that this negative statement, known as the null hypothesis”  isn’t compatible with the data. That is what is meant by “disconfirming the null hypothesis.”

        After gathering all the data on the subject, the researchers use statistical methods to test whether they can reject the null hypothesis. In traditional scientific terms, they conclude they can reject the null hypothesis only if they are at least 95% certain that the null hypothesis is inconsistent with what the data actually show. That is, they look to see if there is only a 5% chance that the findings from their experiments—in this case that carbon dioxide emissions really have increased since the introduction of the gas-burning engine—is just a fluke and wouldn’t be found again if they repeated the experiment.

        The reason scientists use this convoluted way of getting at the truth is because it really is impossible to be 100% certain about almost anything. You may think it is 100% certain that the sun will rise tomorrow, but that prediction is based on the fact that the data from years of experience show it has risen every day for as long as humans have been able to make that observation. There is a tiny chance that tomorrow could just be the day that something cosmically unpredictable occurs, the earth is knocked out of its orbit, and the sun doesn’t rise. And if we cannot be 100% sure the sun will rise tomorrow, what right do scientists have to say anything they find from their experiments is absolutely true?

        Thus, the very methodology that scientists use to find the truth is based on the assumption that nothing is ever completely, 100% certain. And that makes scientists very uncomfortable telling non-scientists they are absolutely sure about anything.

Almost Sure Opens the Denialists Door

        As you can see so clearly from the dialogue that Rich so beautifully captured, scientists at the conference balked at saying they had absolutely zero doubt that climate change is a real phenomenon, even though none of them really had any doubt. And whenever scientists say something like “very sure” or “very likely” or “almost certain,” the door opens for non-scientists to misinterpret that cautious approach to mean that there is real room for doubt, that the science is not yet settled, and that no action need be taken until further experiments are performed.

        In fact, climate scientists are sure that the earth is warming beyond sustainable levels due to our burning fossil fuels and, to a lesser but still significant extent, raising beef. As the American Association for the Advancement of Science (AAAS) put it back in 2014 “The evidence is overwhelming”. Yet put actual scientists in a room with policymakers who are skeptical, and they begin to waffle, using those tell-tale phrases like “almost sure,” and “pretty certain.”  As author Andrew Winston pointed out, it is often not the people go so far as to deny the entire basic science of climate change that we have to worry about; rather, it is what he calls “action deniers,” people who grudgingly admit that climate change is happening because of human activity but insist that not enough is known about it to justify taking any action yet . And these action deniers often justify that position by citing the lack of “we are 100% certain” statements from climate scientists.

        To get a sense of scientists’ discomfort with dramatic descriptions of their work we can look at their reaction to an article in New York Magazine published last year called “The Uninhabitable Earth,” by David Wallace-Wells. Using headlines like “doomsday,” “heat death,” and “the end of food,” Wallace-Wells paints a totally bleak picture of unmitigated destruction and despair. But scientists were unhappy with what they felt to be his hyperbolic presentation. “Instead of welcoming [David Wallace Wells] as an ally,” wrote Genevieve Guenther in Medium, “some climate scientists attack him, roundly criticizing his article for supposedly inspiring paralyzing sense of doom in its readers and, more importantly, for lacking scientific credibility”. In fact, Guenther argues, “The Uninhabitable Earth” was not an exaggeration of the consequences of climate change, but an accurate exposition of them, only written in the prose used by writers but not by scientists.

        There are some signs that scientists are becoming more assertive. In an article published in the peer-reviewed journal Bioscience last April, fourteen scientists took on climate denialists who claim that polar bears are not endangered by climate change. They specifically criticized a blog called “Polar Bear Science”, run by Susan J. Crockford, whom the researchers say  “has no demonstrated expertise in climate science or its effects on polar bears”. This is a somewhat unusual instance in which scientists with clearly relevant credentials push back against denialists, using strong language that indicates they are sure of their facts.

        It is not that we want scientists to violate the terms of the scientific method and start making boldly absolute claims for what their research shows. When scientists become overconfident in their findings, mistakes happen. We want researchers to be cautious, always acknowledging that theirs is a world in which facts do indeed continuously changed whenever new data emerge that demand it. And yet there are things about which scientists are certain. No astronomer today doubts that the universe is expanding. All biologists agree the evolution is the mechanism by which species, including ours, are created. Medical scientists state without reservation that cigarette smoking causes cancer.

        What is the best way for scientists to express certainty without having to say they are certain? Genevieve Guenther in her Medium article believes it will take a division of labor. “We must let the writers strike, pierce, and possess the sight of our souls with what may lie before us and our children if we do not rise up and fight. It is not the role of the scientists to do.”  She may be right. If scientists are uncomfortable stating in front of Congressional panels that there is absolutely no doubt climate change is happening because of things humans do, then perhaps we must turn that job over to scientist-guided writers who have no such qualms. Right now, we need to try every avenue possible to get the message across to the public. Not to be hyperbolic– we are scientists after all–but our very lives depend on it.

More Like This