When Climate Communication Studies Start Sounding Like Self-Help Manuals for Technocrats

Every so often, Nature Climate Change publishes something that reads less like a scientific analysis and more like a frustrated internal memo from technocrats who can’t believe the general population hasn’t fallen in line yet. The latest installment, “Negative verbal probabilities undermine communication of climate science,” belongs to an increasingly familiar genre: studies attempting to psychoanalyze the public for failing to respond properly to elite messaging.

The central premise of this particular paper is as follows:

“We suggest that the negative verbal probabilities prescribed by the Intergovernmental Panel on Climate Change (IPCC) to communicate low-probability outcomes might inadvertently sow doubt about their evidential and consensual base.”

This is the sort of sentence that makes seasoned observers wonder whether the authors realize what they are admitting. If the public hears uncertainty and concludes there is uncertainty, the problem—according to the authors—is not that the uncertainty exists, but that people notice it. To solve this, they propose that climate communication should employ more uplifting diction.

Abstract

The Intergovernmental Panel on Climate Change (IPCC) recommends describing low-probability outcomes using negative verbal probability terms such as unlikely, rather than positive terms such as a small probability. However, we propose that this choice of probability terms might undermine public perception and understanding of climate science. Across eight preregistered experiments (N = 4,150), we find that participants perceive outcomes described with negative low probability terms as reflecting lower scientific consensus than probabilistically equivalent but positively framed terms. The effect persists after controlling for beliefs in climate change, familiarity with the IPCC and political orientation, although it weakens when the projected values exceeded participants’ personal expectations. Participants also associate negative low-probability terms more strongly with extreme outcomes and judge them as less evidence-based than their positive counterparts. We recommend using positive verbal probabilities to communicate comparable levels of uncertainty without undermining perceptions of scientific consensus and evidence.

The paper opens with the standard framing about the dangers of uncertainty:

“Climate misinformation exploits the existence of uncertainty in climate science to sow doubt about how much scientists agree on climate science and fuel climate change scepticism.”

Yet what follows is not a demonstration that the underlying uncertainty is small, well-bounded, or properly constrained. Instead, the authors present eight experiments showing that ordinary people often interpret words as they are written.

For instance, when participants were told something was “unlikely,” they tended to think it probably wouldn’t happen. The authors treat this behavior as a cognitive quirk requiring correction:

“The presumed negative verbal probabilities… were mostly explained by con-reasons… whereas the two presumed positive phrases were mostly completed by pro-reasons.”

Put in plain English: say something is unlikely, and people think of reasons why it won’t occur. Say there is “a small possibility,” and they think of reasons it might. This is presented not as the normal operation of human reasoning but as a messaging flaw.

The researchers confess a further worry: negative phrasing seems to reduce the perceived scientific consensus. In multiple experiments, the public seemed to think that when scientists describe low-probability events with negative wording, it reflects disagreement among experts:

“A scientist describing a temperature rise as ‘unlikely’ disagreed with other scientists more often than when a scientist used ‘a small possibility’.”

The fact that people associate uncertainty with disagreement is treated as a crisis rather than a rational deduction.

The study demonstrates an almost touching faith in the power of verbal cosmetics. The authors note with relief that positive phrasing consistently inflates perceived consensus compared to negative phrasing:

“Perceived consensus was higher for positive than for negative verbal probabilities.”

This is not presented as a cautionary finding—e.g., “wording can distort perception”—but as actionable guidance for communicators hoping to maximize the impression of unified scientific authority.

An especially revealing moment occurs when the authors argue for what they call “complementary likely framing.” This involves abandoning statements such as “a high temperature increase is unlikely” and replacing them with “a lower increase is likely.” They proudly observe:

“Examining the sixth IPCC report we find the term ‘likely’ used 26 times more often than ‘unlikely’.”

The authors appear convinced this is evidence of effective communication rather than a strategic preference for phrasing that sounds more confident and less conditional. In fact, they later celebrate how well this inversion works:

“Participants believed that more scientists would agree with the ‘likely’ than the ‘unlikely’ projections despite the formal equivalence of both statements.”

In other fields, deliberately choosing the phrasing that produces the illusion of greater agreement would raise ethical questions. Here, it is recommended.

Their enthusiasm continues when they compare positive verbal framing to numerical probabilities:

“They inferred more consensus for the ‘small probability’ and the numerical probability projections than for the ‘unlikely’ and ‘likely… not’ projections.”

In other words, if you want people to believe something is solidly evidenced, simply avoid the word “unlikely.” The underlying uncertainties don’t have to change—just the vocabulary.

In the discussion section, the authors finally distill the message:

“We recommend using affirmative low-probability expressions… rather than high-probability expressions of outcome complements.”

Put bluntly: use the phrasing that increases the appearance of consensus and evidential strength, regardless of whether the evidence itself is ambiguous.

Perhaps the most inadvertently illuminating sentence in the entire paper is this one:

“Uncertainty should not fuel scepticism.”

This is a remarkable claim. In every other domain of scientific inquiry, uncertainty is precisely what demands skepticism. Suspended judgment is the disciplined response to unknowns. But in climate communication, uncertainty is treated as a messaging obstacle to be managed so it does not interfere with policy agendas.

The public’s reactions documented in this paper—interpreting “unlikely” as unlikely, associating uncertainty with disagreement, expecting solid evidence behind confident statements—are not cognitive errors. They are the marks of people thinking normally. What the authors diagnose as perceptual bias is simply the public refusing to treat probability bands and verbal hedging as evidence of settled truth.

The irony is that while the authors aim to strengthen trust in climate projections, their own methods highlight why many people remain skeptical. When communication strategies are openly tailored to shape impressions of consensus instead of forthrightly communicating the limits of knowledge, the public intuitively understands what is happening.

And no amount of positive phrasing can hide that.