Let me start with a confession. I work in science communication — which means I spend a non-trivial amount of my professional life explaining to people that science is, by nature, uncertain. That uncertainty is not a bug, it’s a feature. That “we don’t know yet” is one of the most honest and productive sentences a researcher can say. I believe this. I teach this.
And then my doctor prescribes me a medication I’ve never heard of, and I go home and read the leaflet from beginning to end, scrutinising every listed side effect — including the ones that affect “fewer than 1 in 100,000 patients” — and I spend the next twenty minutes genuinely torn between taking the pill and risking a rash, or not taking it and risking whatever the medication was supposed to prevent.
I know perfectly well that the side effects listed in a drug leaflet are not predictions. I know they’re disclosures — documented possibilities, not likelihoods. I know that “1 in 100,000” is a vanishingly small probability. And I still hesitate, because somewhere in my very human brain, uncertainty reads as danger, and I want someone to just tell me it’s going to be fine.
I suspect most of you have been in exactly that place.
Which is why what I’m about to argue might feel a little uncomfortable.
Science speaks carefully. The world rewards confidence.
Scientific language is, by design, cautious. Not because scientists are timid people — though some are, and that’s fine — but because the methodology demands it. A well-conducted study doesn’t “prove” things. It produces evidence that is “consistent with” a hypothesis, within a certain “confidence interval,” with a “margin of error” that is clearly stated.
When researchers are being responsible, they say things like “the data suggest,” “our findings are compatible with,” “further research is needed to establish causality.” These phrases are not hedging. They are the vocabulary of intellectual honesty.
The problem is that public debate — political debate, media debate, the kind that happens on television and in comment sections — does not reward intellectual honesty. It rewards clarity, speed, and confidence. A sentence that begins with “the evidence is compatible with” loses the room before it gets to the verb. A sentence that begins with “there is no proof” wins instantly, even if what it’s describing is the same body of research, framed from the other direction.
This is not a minor communication challenge. It is a structural asymmetry that has been exploited, deliberately and systematically, for decades.
Meet the people who weaponised doubt
In 2010, historians of science Naomi Oreskes and Erik M. Conway published Merchants of Doubt, a book that explains one of the most elegant and depressing communication strategies ever invented: if you can’t defeat science, make it look unfinished.
The tobacco industry perfected the method first. By the 1950s, executives already knew cigarettes caused lung cancer. Internally, the science was settled. Publicly, however, they funded studies, experts and endless debates designed not to prove cigarettes were safe, but to create the impression that scientists were “still discussing the issue.”
Doubt itself became the product. Because as long as uncertainty remained alive, regulation could always wait politely outside the room.
What makes the story almost surreal is how reusable the strategy turned out to be. Acid rain, ozone depletion, secondhand smoke, climate change: different crises, same script. Sometimes even the same people. The topic changed; the choreography stayed exactly the same.
And the choreography is remarkably sophisticated. The merchants of doubt rarely shout. They do not storm television studios dressed as villains from a dystopian Netflix series. They sound calm, rational, almost reassuring. They ask for “prudence.” They request “more evidence.” They remind everyone that “the literature is still inconclusive.” Which is precisely why the strategy works so well. It doesn’t look like an attack on science. It looks like science wearing a very expensive suit.
Nobody actually waits for certainty
Here is the part I find genuinely maddening, once you see it you can’t unsee it: the standard of certainty demanded from science before political action is taken is a standard that we apply nowhere else in human life.
Think about mayors receiving an alert at two in the morning that a river is rising and the downstream valley might flood by dawn. The meteorological models have confidence intervals. The historical flood data has gaps. There is no certainty — there is a probability, and a very consequential asymmetry: if they order an evacuation that turns out to be unnecessary, the cost is inconvenience and money. If they don’t order an evacuation and the flood comes, the cost is lives. They doesn’t wait for certainty. They cannot.
Decisions exist precisely in the space where certainty is not available.
This is true of every serious decision we make. Surgeons operate on probabilistic outcome data, not guarantees. Engineers build bridges with safety margins and load calculations, not certainties. Generals act on intelligence reports that are always incomplete. Insurance companies, pharmaceutical regulators, public health agencies — the entire infrastructure of modern life runs on something far more practical than certainty: it runs on sufficient evidence that acting is better than not acting.
Which makes it all the more strange that when it comes to environmental policy, or public health, or climate regulation, we suddenly discover a previously unknown philosophical commitment to waiting for absolute proof. A commitment, by the way, that is curiously selective — it appears reliably whenever action might interfere with a profitable status quo, and disappears just as reliably when the same companies lobby for regulatory approvals of their own products.
The asymmetry we need to talk about
And here is the part that requires a bit of sitting with, because it touches something we don’t love to examine in ourselves.
Many of us — myself included, as I admitted at the start — feel uncomfortable with scientific uncertainty in areas that affect us personally. We want the leaflet to say “this will work and nothing bad will happen.” We want the climate report to say “here is exactly what will occur and exactly when.” We want the pandemic model to say “this is definitely what’s coming.” And when science doesn’t say those things — because it can’t, because that’s not how knowledge works — we feel, irrationally but humanly, let down.
The merchants of doubt understand this feeling intimately. They are, in a very real sense, in the business of exploiting it. They know that if they can keep the focus on what science hasn’t yet proven rather than on what it has, they can maintain the emotional conditions — vague unease, sense of unresolved debate — that make people reluctant to support action. They’re not fighting the science. They’re fighting the psychology.
The reframe that I think we urgently need — and that science communication urgently needs — is a shift from the question “is this certain?” to the question “is this enough?” Because on climate change, to take the most pressing example, the answer to the first question is, technically, no: there are still error margins, still scenarios, still debates about specific regional impacts. And the answer to the second question is: yes, overwhelmingly and for some time now. The evidence is enough to act on. It was enough twenty years ago. Waiting for it to become certainty is not a neutral position — it is a decision, and it has consequences that are, at this point, anything but uncertain.
The mayor at two in the morning knows this. The question is whether the rest of us are willing to catch up with her before the river crests.