The power of not-knowing

While knowledge is important, a position of not-knowing about the threats we face sparks curiosity and the motivation to navigate uncertainty together.

Chemical pollution is, one researcher recently suggested, "a threat to the thriving of humans and nature of a similar order as climate change, but is decades behind global heating in terms of public knowledge." They went on to say, "a lot of people assume that there's really great knowledge and huge due diligence on the chemical safety of these things. But it really isn't the case."

This illustrates how what we know and don't know sits at the centre of some of today's most charged debates about how we deal with societal and planetary challenges. Typically, and perhaps not unreasonably, it is assumed that both experts and the public should have the proper knowledge to identify solutions and drive positive change. This is what makes 'misinformation' such a potent topic, as it is fundamentally a debate concerning what is known and what is not known, what is legitimate and what is illegitimate knowledge. Media literacy and critical thinking are tools to navigate our ‘knowing’ and, as such, are often called upon as part of this information 'warfare'.

If 'knowing' is central to how we manage the big issues of our day, then what are the implications for policymakers and marketers who hope to influence change in behaviour? At a basic level, it makes sense to us that to know something is a prerequisite for taking action. But, as we shall see, there is also a great deal of evidence that whilst simply giving people information (so they 'know') is important, it is far from sufficient.

Our pathway through this is to make the counter-intuitive case that 'not-knowing' can, in fact, be much more important and at times more helpful than 'knowing'. And whilst there is much researched and theorised about knowledge and 'knowing,' much less is explored and written about being in a state of 'not-knowing,' arguably a much more familiar human condition.

We argue that being willing to accept states of not-knowing is not only inevitable but also something to welcome and support if we are to address the complex challenges we now face, such as chemical pollution.

The case for knowing

But before exploring the power of not-knowing, it is worth acknowledging why knowing, or at least the perception of knowing, matters so much in the behavioural science literature.

One of the clearest cases is made through what is known as the Competence Hypothesis. This body of research suggests that people are more likely to take action when they feel knowledgeable in a particular domain, even when that knowledge is superficial or incomplete. In other words, we don't just act when we are competent; we act when we feel competent – a small but significant distinction.

A well-known example comes from a study by Chip Heath and Amos Tversky comparing how likely sports fans are to place a bet on a match versus a random lottery. Despite the statistical odds being similar, the fans were far more comfortable betting on sports. This was due to how they knew the teams, or at least felt they did, as they had watched the games, recognised the players, and heard the commentary. This domain familiarity gave them perceived competence, which in turn reduced perceived risk, increasing their willingness to act in a way they did not with the lottery. Hence, this is not simply about statistical reasoning but about affective trust in one's own judgement.

We also see this echoed in other behavioural domains. For example, people are more likely to invest in companies they have heard of, regardless of whether that knowledge is meaningful. And this is why brand familiarity has long been recognised as a potent factor in decision-making. On this basis, the act of knowing doesn't just inform the decision; it legitimises it, giving people a sense of control, even if the control is ultimately illusory.

In behaviour change interventions, this matters. When people feel equipped, such as by understanding the wording, seeing examples they relate to, and having prior exposure to similar scenarios, they are more likely to engage. Whether quitting smoking, changing diets, or switching to renewable energy, a sense of competence makes the leap feel safer.

So while we will be making an argument for the power of not-knowing, for now, we start with this acknowledgement: in many situations, perceived knowledge increases behavioural willingness. It makes action feel less like a gamble and more like an informed choice.

But knowing does not always work

It is generally understood that while improved knowledge is often considered necessary for good decision-making, it is far from sufficient, as simply giving people information does not guarantee that they will act on it.

The scale of this discrepancy is illustrated by a well-known meta-analysis on financial behaviour by Daniel Fernandes, John Lynch, and Richard Netemeyer examining financial education programmes. They found that these programmes only explained 0.1% of the variance in financial behaviours, a vanishingly small effect. So, despite the widespread belief that better literacy should lead to better decisions, in financial behaviours at least, there is significant evidence that knowledge alone does not always translate into action.

Part of the issue here is that behavioural interventions all too often assume people act as purely economic actors, so that if we are well enough equipped, we will make good choices. But, of course, daily life rarely plays out that way - we are busy, distracted and overwhelmed. Making choices in a complex, seldom-visited financial market requires time and mental effort, which are often in short supply when managing relationships, work, bills, or just trying to get through the day.

And even when knowledge is delivered, it does not always stick. The paper's authors show that the effects of financial education tend to decay over time, quickly forgotten unless anchored to a real, immediate decision. That's why they advocate for ‘just-in-time’ education: targeted interventions delivered at the exact moment a decision is being made. Timing, it seems, and not just content, matters.

Other promising approaches shift the focus from abstract knowledge to practical relevance. Economists Alejandro Drexler, Greg Fischer, and Antoinette Schoar tested this in the Dominican Republic, working with small businesses. Rather than offering textbook financial training, the team distilled simple, context-relevant 'rules of thumb' drawn from the habits of successful local enterprises. These heuristics were short, concrete, and directly tied to everyday business needs: those who received this rule-of-thumb training saw sales increase by up to 25%. In contrast, those who received standard financial literacy training saw little change.

This supports a broader point: people don't just need to know more, they need to feel that what they know fits into the rhythm of their lives. Knowledge that is not readily understandable, relatable, or actionable can quickly fade.

But what we might call this 'limits-of-knowing problem' is not just about attention or relevance; it's also about appetite. Even when knowledge is delivered effectively, it can demand more from people than they can offer. This is the Cassandra effect: when knowing too much does not spur action but instead shuts it down.

Not wanting to know too much: The Cassandra problem

In Greek mythology, Cassandra was cursed with the gift of true prophecy, but no one believed her. Today, the 'Cassandra effect' has become a modern parable: not just about being ignored, but about the emotional toll of seeing what is coming and being powerless to stop it.

Gerd Gigerenzer and Rocio Garcia-Retamero developed a psychological underpinning to this, arguing that people often avoid foreknowledge not out of irrationality, but to shield themselves from anticipatory regret. Using a nationally representative survey across Germany and Spain, they showed that deliberate ignorance is widespread, finding that between 85–90% of respondents preferred not to know about impending adverse events (such as the time of their partner's death), and 40–70% even avoided foreknowledge of positive outcomes (such as the sex of a child).

The study also found that people are more likely to avoid knowledge as the anticipated event draws nearer, because the potential for regret intensifies. Hence, 'not-knowing' often functions as a rational strategy for preserving dignity, hope, and the possibility of joy in the present by refusing to let foreknowledge impact current lived experience.

This very issue was illustrated in the film The Farewell, directed by Lulu Wang. Based on a true story, the film follows a Chinese-American woman who returns to China to visit her grandmother, Nai Nai, who has been diagnosed with terminal cancer. But Nai Nai doesn't know; her family has chosen not to tell her, instead gathering under the guise of a wedding to say goodbye. The decision reflects a practice common across East and South Asia: shielding loved ones from harsh truths not to deceive, but to protect. For Nai Nai, not-knowing becomes a kind of mercy, allowing her to continue living with dignity and joy, unburdened by fear.

And deciding to 'not-know' is widespread. A Reuters Institute study across 17 countries recently found that 39% of people actively avoid the news, up 10% in less than a decade. For many, this is not about disinterest but protection against emotional fatigue, especially visible around environmental concerns. For example, 69% of Gen Z feel anxious after encountering climate content online, far higher than older generations, reinforcing the point that people may reject information not because they disbelieve it, but because they cannot sustain the emotional weight of knowing.

For policymakers and communicators, it is therefore very helpful to recognise that more information is not always better. What may well matter more, in fact, is whether people feel supported and able to act without being overwhelmed.

How risks are becoming less knowable

Adding to the way that not-knowing can act as an adaptive strategy, is that way that the risks we face are becoming increasingly less knowable than in the past. Ulrich Beck writes about this in the Risk Society: he suggests that the science, industry, and technology that have been built to manage old dangers have created new ones, which often exceed comprehension. How we then deal with them is also less 'knowable' as they now cross borders, generations, and systems, making them indeterminate and unmanageable through traditional means.

Chemicals appear to embody this shift from old dangers to new. They were first hailed as solutions: fertilisers to tackle famine, pesticides to safeguard crops, chlorine to make water drinkable, flame retardants to prevent household fires. Each was designed to shield humanity from threats that had haunted earlier generations, such as hunger, disease, accident, and decay. And they have been hugely successful in these missions.

But the very tools that solved yesterday's challenges all too often seem to generate new dangers of their own. Fertilisers now leach into rivers, creating dead zones; pesticides accumulate through ecosystems, destabilising species and human health; chlorine derivatives linger as toxic by-products; flame retardants are found in human blood and breast milk.

It’s well known that the industrial economy has created millions of novel chemical entities, and yet we lack the tools to trace their long-term effects. Endocrine-disrupting chemicals, for instance, can produce more potent effects at low doses than at high ones, a pattern that defies traditional toxicology models. Here, the risk is not just considerable; it is structurally resistant to being known.

The uncertainty does not end here either. Sociologist Anthony Giddens wrote that we now live in a condition of ‘reflexive doubt’, where established sources of authority and expertise are continuously questioned, and knowledge is subject to constant scrutiny and revision. Climate scientists update projections with every new dataset; economists re-run forecasts in the face of unpredictable shocks; epidemiologists adjust guidance as viruses mutate. And the public, increasingly aware of these revisions, is less willing to treat expert pronouncements as settled fact but as provisional, subject to correction, contradiction, or reversal.

Artificial intelligence also plays into this atmosphere of unknowability. Algorithmic decision-making introduces risks that are not only opaque to laypeople but often to the specialists who build or regulate these systems. Black-box models can deliver outputs that are difficult to fully explain, leading to consequences such as biased hiring systems and misfiring credit scores, which often go undetected until after the harm is done.

All this suggests that the challenge is not simply that risks are multiplying, but that they are becoming resistant to complete comprehension. They defy the older dream of mastery through calculation and instead demand that we navigate ‘not-knowing’ as a permanent condition.

Institutions of knowing could do more to help navigate uncertainty.

Although the world may be less knowable, some critics suggest that our institutions all too often push in the opposite direction. These critical voices suggest that schools, universities, and workplaces can be reluctant to acknowledge uncertainty, all the while performing certainty. Success, it is argued, is measured by confidence, fluency, and decisiveness, even when the underlying knowledge is partial or fragile. The result is a culture with little room for doubt or hesitation.

The university, long imagined as a place of exploration and doubt, has come in for particularly sharp criticism. Scholars such as Bill Readings (The University in Ruins) and Martha Nussbaum (Not for Profit) argue that universities have shifted from cultivating open inquiry to producing measurable outputs such as rankings, impact metrics, and employability scores. This is not just bureaucratic drift but a profound epistemic narrowing. As the theorist Stanley Fish once put it, universities increasingly reward "being right" over "learning to think."

The state of knowledge reflects what William Deresiewicz calls "world-class hoop jumping": signalling authority or inferring there is a solid position even where the knowledge is fragile. In his book Excellent Sheep, he describes how elite students are trained to be 'overachieving but under-inquiring': brilliant at fulfilling expectations, but rarely asking why those expectations exist. On this basis, not-knowing quickly becomes a marginalised activity - economically punished, socially pathologised, and pedagogically erased. Knowing becomes less a process of discovery (which inevitably involves not-knowing) than an identity, a signal of belonging.

In a Risk Society defined by reflexive doubt, this is a dangerous irony: critics suggest that the very places that should surely teach us to navigate uncertainty, instead run the danger of training us to deny it.

We live in the not-knowing (whether or not we know that!)

Perhaps most importantly, philosophers have long pointed out that not-knowing is core to the human condition, the ground we all stand on. In fact, societies only function because we live with, and through, what we don’t know (rather than what we individually know). Philosopher of language Paul Grice showed that communication rests less on fact-checking of knowledge than on co-operation, we assume sincerity, and we trust without verifying the facts. As psychologist David Dunning (of Dunning-Kruger effect fame) has argued, without this tacit willingness to take others at their word, everyday life would be unworkable. We don't interrogate every claim from friends, colleagues, or strangers; we have to inhabit a world where some ignorance is not only tolerated but necessary.

This extends to how, despite this state of not-knowing, we think we know. As we have often pointed out, psychologists Steven Sloman and Philip Fernbach have demonstrated that we all operate with a knowledge illusion: that people routinely believe they know more than they do, because they intuitively outsource detail to their community. We can say 'mortgage,' ‘arthritis,' or 'algorithmic bias' without being able to explain them entirely, because the social world provides scaffolding that makes our individual shallow knowing usable. Hilary Putnam called this the division of linguistic labour: words and labels carry the weight of communal expertise, even if no individual holds it all. In this sense, our not-knowing is not a bug but a design feature.

Philosopher Ludwig Wittgenstein famously understood this when he insisted that meaning is not held privately in our heads but in shared use. A label is meaningful because others can agree or reject it; its explanatory force comes from communal entrenchment. Knowing, then, is less about mastery than about being able to move fluently in a system where the bulk of knowledge is elsewhere.

Seen this way, not-knowing is not an obstacle to social life; instead, it is its very medium. We survive, coordinate, and build futures not by erasing ignorance but by weaving it into a collective fabric, as the demand that every individual always know is unrealistic. Our capacity to live together depends on accepting that much of what sustains us, from trust to expertise to shared labels, rests on things we do not, and cannot, fully know.

It seems odd then that we fetishise knowing when in fact our more common state is not knowing - and likely becoming more so as the risks we face become more unknowable.

Curiosity is an important characteristic of not-knowing

One very helpful human characteristic of not-knowing is that it can spark an essential human characteristic - that of curiosity. We see this reflected in Nick Chater and George Loewenstein's paper, where they set out the case that the curiosity that comes from not-knowing is not a luxury but a drive (like hunger or thirst) borne of where we notice an information gap and feel compelled to bridge it. And if the gap is closed too quickly, then the drive dissipates. So whilst having a sense of certainty satisfies in the short term, it risks shutting down the energy that fuels exploration and learning. In other words, premature 'knowing' can act as a 'curiosity kill switch'.

By contrast, leaving some questions open sustains the drive to make sense. Behavioural research suggests that grappling with competing cues and partial truths, rather than being fed neat answers, can result in longer-lasting engagement and deeper understanding. The discomfort of ambiguity is not an obstacle to change but the ignition.

And linked to this is the collective nature of not knowing: if the problems we face are full of collective uncertainty, then curiosity also becomes collective, a resource we pool and protect. Citizens' climate assemblies across Europe have shown what this looks like: when ordinary people are invited to grapple with messy trade-offs, rather than being handed neat conclusions, the result can be deeper trust, greater legitimacy, and more robust recommendations. Another example of this is from Ireland, where assemblies on abortion and same-sex marriage created the conditions for historic constitutional change because citizens were treated as capable of holding uncertainty together.

And during the COVID-19 pandemic, evidence suggests that countries that openly acknowledged uncertainty and explained what was known and unknown maintained higher public trust than those that insisted on false clarity. In the current heated debates over AI governance, we see something similar: the most credible conversations seem to be not those promising 'safe AI' through technical fixes alone, but those that admit the limits of knowledge and invite citizens, workers, and affected communities into ongoing negotiation.

It may be time to rehabilitate 'not-knowing', as there is a strong case that it is not the opposite of progress but its very condition. It legitimises hesitation, sustains contestation, and keeps questions alive. It creates the space for course correction and collective learning. Which is why rushing to foreclose uncertainty, to insist on the performance of certainty, doesn't just flatten curiosity but undermines our ability to act together at all.

The politics of knowing versus not-knowing

One important consideration here is, of course, that we do not live in a world where all parties might agree to 'not-knowing' and curiously respectfully debate the points!

For example, on the topic of chemical pollution, consider the recent debate about synthetic food dyes, fuelled by wellness influencers and politicians such as RFK Jr. The claim that dyes like Red 40 are 'derived from petroleum' has been made as though it were evidence that children were "eating gasoline." In reality, as immunologist Andrea Love points out, this is a category mistake: petroleum-derived building blocks are transformed into entirely new molecules through chemical processes, just as bread is not "wheat stalks" and salt is not explosive sodium.

This is an example of the debate being less about the chemistry and more about the performance and politics of knowing - what we might call 'inflated knowing'. Love suggests that these sorts of claims work rhetorically because they create two camps: those who are 'awake' to (and ‘know’ about) hidden dangers and those who are 'sheep’ and ‘don’t know’.) In that sense, Love argues, some actors can also monetise these concerns by selling supplements, ‘natural’ dyes, or lifestyle hacks, dramatising knowledge as something withheld by 'corrupt experts' and revealed by 'brave truth-tellers.'

Conversely, (and equally unhelpful but from the other direction) is the taking of a position of wilful ignorance (or as we might call it for our purposes, ‘wilful not-knowing'). One recent meta-analysis found 40% of participants avoid easily obtainable information about the consequences of their actions on others, leading to a 15.6-percentage point decrease in altruistic behaviour compared to when information is provided. In chemical risk, there are claims that strategic ignorance has historically been incentivised by fragmented testing regimes and liability design, making the absence of evidence look like evidence of absence.

In conclusion

The implications for those involved in behaviour change suggest a delicate balance between knowing and not-knowing. It would be a foolhardy act to design a behaviour change programme without recognising the importance of information, albeit delivered in a way that is relevant, timely and consistent with the way the target population understands the world.

But to leave it at this means we risk acting in a manner that is too 'top-down', appearing to dispense explanations (with accompanying guidance on what to do) to an uninformed population who 'ought to know better'. This approach is rapidly looking outdated; people want to learn from each other, be 'part of the solution'. Curiosity is what researcher Carmen Valor might call an 'activating emotion', something that encourages behaviour to change.

Of course, this creates a challenge for those who represent bodies of expertise – how to navigate the knowing versus not-knowing. Someone coming in for advice on a medical condition may not want to be made aware of the limits of medical knowledge, but others might definitely want to. However, some actors may also attempt to undermine bodies of expertise by claiming what we don't know, when in fact these are issues with firm bodies of evidence to support them. There is a need for experts to develop the necessary skills to navigate this rapidly changing epistemic environment.

But, despite the complexities, there is a good case for a greater openness about what we know and don't know. While governments, brands and regulators may at times be uncomfortable taking a more nuanced position on issues such as chemical pollution (as there is then an opportunity for some actors to take advantage of this nuance), it does seem that a shift is inevitable as people look for information and guidance from more horizontal sources (e.g. such as social media).

Overall, not-knowing is the condition under which we have always lived and will continue to do so (perhaps even more so). While much critical work goes into mastering what we know, we surely also need to focus more on how we can better navigate not-knowing together.