Two years ago at the March for Science in Boston, I saw signs that said "Keep Calm and Think Critically," "Extremely Mad Scientist," "No Science, No Twitter" and "It's So Severe, The Nerds Are Here." If anything, since then things have gotten worse.

Virtually every day we find an assault on science in the pages of our newspapers and on TV, the radio and the Internet. Seven hundred measles cases in 22 states due to anti-vaxxers. Climate change legislation stalled in Washington, D.C., due to willful ignorance over the difference between climate and weather. Even the flat-earth movement is on the rise.

The attacks target not only the findings of science, but the very process by which well-corroborated scientific theories are produced. Perhaps few if any of the lay public ever really understood how science works, but in decades past they at least trusted scientists as experts. In today's hyperpoliticized, do-your-own-research-on-Google era, a good deal of that trust has vanished.

In a recent survey cited in Scientific American, only 35 percent of respondents had "a lot" of trust in scientists, while the number of people who "do not at all" trust scientists increased by over 50 percent from 2013 to 2017.

The question of what accounts for the special epistemic authority of science is no longer purely academic. If we cannot do a better job of defending science—of saying how it works and why its findings have a privileged claim to believability—we can only expect things to get worse.

In my new book The Scientific Attitude: Defending Science from Denial, Fraud, and Pseudoscience (MIT Press, 2019), I propose that the best way to defend science is to reconceive what is most special about it. This entails abandoning several longstanding myths about science that have given aid and comfort to deniers.

In brief, my claim is that what is most distinctive about science is not any "scientific method" or "logic of science" that philosophers and others have tried to use as a criterion of demarcation between science and nonscience. Instead, the key feature is the "scientific attitude," which consists of two theses: (1) scientists care about evidence and (2) they are willing to change their beliefs based on new evidence. Rather than any method or logic, the scientific attitude is an essential value of science. It is embodied in the scientific community’s practice of testing and checking one another's work, without which science could not move forward.

At this point, some of you who regularly read these pages might wonder, "Why do philosophers always think we need help from them?" Perhaps the most diplomatic answer is that this post-truth era is not the time to refuse help from your allies. A more provocative answer is to turn the question around: How have things been going in your own defense of science to a lay public who do not really understand what scientists do?

The typical way that scientists defend science is to present their evidence and then, if it is not rationally evaluated or their integrity is questioned, walk away. Perhaps it is understandable that people who have devoted their lives to rigorously testing empirical theories would become exasperated with faux-skepticism or conspiracy theories and would dismiss these ideas as irrational.

Indeed they are irrational. But at a time when even the most patent nonsense gets amplified on the Internet—sometimes reaching all the way to Congress and the White House—I implore scientists not to walk away (nor merely to protest), but to engage in more public education not just about the results of science but about how science works. Remember that every lie has an audience. If we do not fight back, we lose the chance not just to try to convince the liar, which may be impossible, but more importantly to provide a counter-argument to the liar's audience.

How, then, to proceed? You cannot convince someone who does not believe in evidence by showing them more evidence. You succeed, if at all, by showing them that their reasoning is faulty. And one of the most important things to do here is to dispel the myth that science, like its cousins math and logic, can produce "proofs"—or that the point of scientific inquiry is to reach certainty.

Unfortunately, when backed into a corner by some know-nothing who claims that climate change "isn't settled science" or that evolution "is just a theory," scientists themselves have given oxygen to these myths. The desire to blurt out "But it's been proven!" can be almost irresistible.

Why do scientists sometimes do this? Because they know that if they admit any uncertainty it will be exploited by science deniers to claim that "until the rest of the evidence comes in" their own theory is just as likely to be true. Of course, this is a ridiculous standard. Just because certainty cannot be reached does not mean that likelihood and probability go out the window.

When certainty is the standard, science deniers may feel justified in calling themselves skeptics and retreating into their weird silos of confirmation bias and disinformation, where any slim possibility can seem like victory. So let's show them that this is an unreasonable standard. Let's show them, at least, that this is not the way scientists reason based on evidence.

I wish scientists would stop being so embarrassed by uncertainty and instead embrace it as a strength rather than a weakness of scientific reasoning. The scientific attitude allows us to be open to new ideas, even while we must remain skeptical of them until they have been thoroughly tested. That is a model of true skepticism.

One virtue of embracing the scientific attitude is that it would allow scientists to say what in their hearts they know to be true: that science cannot, technically speaking, ever prove a theory true—even that gravity exists or that electrons are real. That is just not how inductive reasoning works. In studying the empirical world we must always hold out the possibility that some future evidence may come along to refute even the best-corroborated theories. Just as Newton’s theories gave way to Einstein's, every scientific theory is fallible.

But in the next breath, we must shut down the false idea that uncertainty allows any wild-eyed theory to be believable. The strength of scientific explanation is based on warrant—on justification given the evidence—and although this may not be a matter of proof, it is a matter of (overwhelming) likelihood and probability. While it is true that, until it has been disproven, any theory may be correct, this does not mean it is rational to believe a theory with insufficient evidence.

Another way to proceed might be to pick a case of uncertainty and talk about that. For instance, it was recently reported that the evidence for anthropocentric climate change has now reached the "five sigma" level. This is the gold standard of corroboration, indicating there is now only a one-in-a-million chance that the climate change deniers are right. Perhaps this might feel like enough for them. But should it?

In the Jim Carrey movie Dumb and Dumber there is a scene where the protagonist is trying everything he can to get a woman to go out with him. No matter how many different ways he asks, she just keeps rejecting him. Finally, in exasperation, he asks her to assess the probability that she would ever date him, to which she replies, "One in a million." This film being the comedic gem that it is, he grins and says, "So you're tellin' me there's a chance."

Instead of supporting myths about certainty, let's make the science deniers own that.