Risk is never neutral (and neither are we)

This reflection grew out of the closing moments of a panel on risk communication that I recently moderated. Different disciplines were represented on stage, and as each speaker described “risk,” it slowly became clear that we were not just offering different perspectives — we were, in fact, speaking slightly different languages.

When we talk about risk communication, the first piece of advice is almost always the same: know your audience. And yes, that matters. It matters a lot. But listening to that conversation, I realised that something else matters just as much — and we tend to overlook it.

Not only who we are speaking to.
But from where we are speaking.

Risk is not a neutral word. It never has been. The way we define it, frame it, and explain it depends on the discipline we come from, the problems we are trained to solve, and the conceptual tools we use every day — often so routinely that we no longer notice them.

I am an economist by training. And that inevitably shows.

When Risk Becomes a Question of Costs

As an economist, I instinctively approach risk as a problem of choice under uncertainty.

The first questions that come to mind are rarely “Is this dangerous?” but rather: What are the alternatives? What are the trade-offs? Who bears the costs, and who captures the benefits?

Economics tends to translate risk into decisions. Not because everything can — or should — be reduced to numbers, but because collective choices require some way of comparing options. Risk becomes part of a broader calculus: acting versus not acting, regulating versus leaving things unchanged, investing now versus postponing.

Time plays a central role. A consequence today does not weigh the same as a consequence tomorrow. Future outcomes are evaluated, implicitly or explicitly, against present costs. That evaluation may look technical, but it always contains a value judgment about how much the future counts.

Probability, in this framework, is rarely a single prediction. It is a distribution of possible outcomes, some more likely, some less so, some unlikely but potentially severe. Economists are particularly attentive to how these different scenarios affect incentives and behaviour.

From this perspective, risk is not only about exposure to harm. It is about how individuals and institutions respond to uncertainty, how they allocate resources, and how they structure responsibility.

In short, risk becomes inseparable from decision-making itself.

A Shift in Language: Probability, Possibility, and Perception

Move into scientific research more broadly, and the vocabulary shifts — although not always in the way it is perceived from the outside.

Within STEM disciplines, probabilities, uncertainties, and confidence intervals are integral to everyday reasoning. Scientists do not choose “possibility” over “probability”; they work with both. The shift happens when this language leaves the laboratory and enters public discourse.

Outside expert communities, people tend to react much more strongly to possibility than to probability. The dominant question becomes not “How likely is this?” but “Could this happen to me at all?”

Possibility feels concrete and personal. Probability feels abstract and statistical. From an everyday perspective, either there is a risk hanging over me or there is not. The gap here is not in the data themselves; it lies in how different audiences cognitively and emotionally process them.

Medicine introduces another variation.

Risk in the Consultation Room

In medical contexts, risk ceases to be about abstract models or territories. It becomes relational and embodied. There is no such thing as zero risk; there is only risk in comparison to alternatives. Treatment versus no treatment. Side effects versus disease progression. Absolute risk versus relative risk.

Moreover, risk is not uniform. It varies with age, medical history, comorbidities, and social context. A percentage does not mean the same thing for every body.

For this reason, medical risk communication cannot be purely unidirectional. It requires dialogue, negotiation, and shared decision-making. Clinicians may speak in probabilities, and they may choose whether to frame those probabilities in absolute or relative terms, but the numbers acquire meaning only within a relationship of trust.

This dynamic became especially visible during Covid. At times, scientists appeared to disagree publicly. An immunologist and a virologist could offer what sounded like conflicting assessments of the same issue. Yet they were not necessarily contradicting each other. They were often focusing on different mechanisms, different temporal scales, different layers of the problem. They were asking different questions of the same dataset.

Within science, such plurality is normal. Outside it, the coexistence of multiple disciplinary lenses can easily be perceived as confusion.

Hazard, Uncertainty, and Responsibility

In fields such as volcanology, the conceptual shift is even more pronounced.

A volcanologist will often begin by distinguishing hazard from risk. Hazard refers to the physical phenomenon itself — what the volcano is capable of doing. Risk emerges only when exposure and vulnerability are added to the equation.

The language used is precise and cautious. Researchers speak of scenarios rather than predictions, conditional probabilities rather than certainties, thresholds and alert levels rather than definitive outcomes. Uncertainty is not treated as a weakness of knowledge but as a structural characteristic of complex systems.

This explains why volcanologists avoid terms such as safe, impossible, or inevitable. Not because they lack information, but because they are acutely aware of how easily such words can be misunderstood once detached from their scientific context.

From this perspective, risk is neither a single number nor a prediction of what will happen. It is an evolving relationship between a dynamic natural system, imperfect monitoring tools, and communities that must make decisions under incomplete information.

What counts as methodological prudence within the discipline may sound, to those living on the territory, like hesitation or ambiguity.

Distance and the Shape of Climate Risk

Climate science introduces yet another layer: distance.

For many people, climate change remains spatially and temporally distant. Melting ice caps, long-term projections, modelled temperature increases by 2100. The scientific language of models, intervals, and scenarios is entirely appropriate within the discipline.

Yet public reasoning rarely unfolds in those terms. The question is often how much this concerns me here and now. When risk becomes proximate — when floods affect one’s region, heatwaves disrupt daily routines, or water scarcity impacts local economies — the perception shifts dramatically. The underlying science may be unchanged; what changes is proximity.

Geographical, emotional, and biographical distance all influence how risk is experienced. Once again, the gap lies less in the evidence than in the interpretive framework through which it is received.

The Semantic Layer: Words That Carry Histories

Beyond disciplinary differences, there is also the question of words themselves.

Some terms arrive already loaded with history. “Nuclear,” for instance, carries decades of imagery and collective memory. The word alone activates a dense semantic landscape before any data enter the discussion.

Replace it with “advanced fission energy,” “fourth-generation reactors,” or “low-carbon baseload technologies,” and the technical referent may remain similar, but the perceived risk shifts. The semantic field changes.

This is not necessarily manipulation. It is a reminder that language shapes perception. Every lexical choice opens a particular cognitive and emotional pathway. There is no neutral vocabulary for risk, because words themselves are embedded in cultural memory.

Communicating risk therefore involves not only explaining data but also being aware of the semantic worlds that our words evoke.

Risk as Meaning, Trust, and Inequality

A sociological perspective pushes the reflection further.

Risk does not exist as a purely natural object. It acquires meaning within social contexts. This does not imply that hazards are invented; rather, what counts as risky and how it is prioritised depends on cultural norms, historical experiences, institutional trust, and power structures.

Perception matters because behaviour is shaped less by measured risk than by perceived risk. Trust matters because data are filtered through judgments about those who communicate them. Inequality matters because exposure and resilience are unevenly distributed.

The same hazard does not have the same meaning in every community. Nor do all groups have equal influence in defining which risks are acceptable.

From this perspective, communicating risk is not simply a matter of clarifying probabilities. It is an intervention in a broader social relationship between knowledge, trust, and power.

The Necessary Negotiation

Across these perspectives — economic, scientific, medical, volcanological, sociological — one conclusion gradually emerges: risk does not exist independently of the frameworks through which it is interpreted.

Distance alters perception. Language reshapes meaning. Disciplinary lenses filter what counts as relevant. Emotions and trust enter long before numbers have completed their work.

Risk, in other words, does not travel intact from one domain to another. It changes shape as it moves between expert communities and the public sphere.

For this reason, communicating risk is not merely a technical exercise in translation. It is a negotiation between forms of expertise, lived experience, and cultural context. Accepting that negotiation is not a dilution of science; it is part of its social responsibility.

And perhaps the next time we are advised to “know our audience,” we might also ask a complementary question: from which intellectual and disciplinary position are we speaking — and how does that position already shape the risk we think we are describing?