This article was written with Nicolas Marescaux, originally in French.

Today, our modern life relies largely on numbers. They guide most collective decisions and many individual choices. For Lord Kelvin [1], “If you cannot measure it, you cannot improve it.” In other words, to make a good decision, you must first measure well. But is that enough? IPCC reports have been compiling data and figures for decades, announcing a short-term catastrophe. And yet, nothing happens. “The modern man scorns imagination,” stated Stéphane Mallarmé in 1897. Isn’t it this subjectivity of our imagination that could save us?

From story to “storytelling”

The ability to count well has been important for organizing life in society and for accessing modernity. Censuses, for example, are foundational acts of many cultures, including Western society. In the Sinai desert, God said to Moses: “Take a census of the entire assembly of the Israelites according to their clans,” as we read at the beginning of the fourth book of the Bible [2]. The census is the basis of demographic organization, whether agrarian, military, or urban, and it also precedes calculation, collection, and taxation—essential elements of collective life. The advent of “Big Data,” as described by Mayer-Schönberger & Cukier (2014), reminds us of the importance of this new “black gold.” Even though numbers are important, they can never capture the full complexity and humanity of our experiences and decisions.

Almost fifty years ago, Roland Barthes noted that there is no people without stories, whatever form they may take: myths, legends, fables, tales, novellas, novels, films, etc. This observation led to the foundation of what would become “narratology,” the idea that storytelling is one of the main categories of knowledge we use to understand and organize the world. A story is a universal fact of language, and its meaning cannot be reduced to the sum of its narrative units: “Reading (or listening to) a story is not just moving from one word to another; it’s also moving from one level to another.” Paul Ricoeur took up this idea in The Course of Recognition (2004), emphasizing that great stories marking human history, from Homer to Tolstoy, from Sophocles to Shakespeare, recount universal myths and convey wisdom derived from the accumulated experience of past generations. But more importantly, he observed that “storytelling” follows an opposite path: it imposes artificial narratives on reality, blocks exchanges, and saturates the symbolic space with series and stories. Instead of recounting past experiences, it traces behaviors and directs emotional flows. Lynn Smith reminds us that “storytelling,” which originated from the postmodern literary (mainly academic) movement of the 1960s, spread into broader culture—psychologists, geographers, physicists, and economists rediscovered the power of stories in shaping reality. Economists Deirdre McCloskey and Ariel Rubinstein argued that economics is essentially a narrative discipline, with mathematical formalism being only the translation of that story or a way to test its robustness.

Storytelling has come to rival logical thought in understanding the world. Lynn Smith quotes Paul Costello, founder of the Center for Narrative Studies: “We always knew stories are really powerful. They’ve probably never been treated before as if they mattered in shaping our public and private lives (…) Before, it was always, ‘That’s only a story; give me the facts.’ Now, more people are realizing that stories have real effects that have got to be looked at seriously.” Charles Darwin did not discover the theory of evolution; it had been in the air for decades. But he found a way to tell that theory by describing anecdotes that, pieced together, sketched a general theory. We remember the finches of the Galápagos Islands, whose beaks had different shapes and sizes, adapted to specific diets, or the mutual adaptation between orchids and pollinators, illustrating the perfection of natural selection mechanisms.

These stories almost tend to replace history. “When the legend becomes fact, print the legend,” said John Ford in The Man Who Shot Liberty Valance. This is the idea of the “national narrative,” an idea widely promoted about fifteen years ago by Henri Guaino, then advisor to Nicolas Sarkozy. It is history seen as a story, or when history (a scientific discipline) tends to be replaced by a duty of memory (and is thus eminently political, given the choices made to highlight this or that figure, to celebrate this or that event). This is also the etymological origin of the word “propaganda,” which refers to what must be propagated (beliefs, mysteries, legends of saints, miracle stories, etc.). It is not about transmitting objective knowledge accessible to all through reason, but about persuading people to adhere to hidden truths that belong more to faith than reason.

Kobi Hackenburg and his co-authors show that generative artificial intelligence is progressing in its ability to produce convincing speeches​​. It can be used to amplify the reach of narratives and emotions in decision-making, for instance, by helping to create detailed and emotionally engaging fictional scenarios that make real stakes more tangible for the public and decision-makers. However, this technology can also be used for disinformation. Steven Brill illustrates how disinformation is funded and spread, while Andy Arditi and his co-authors explain how potential directions can be imposed to improperly release a large language model​. John Timmer highlights the role of “super spreaders” of misinformation on Twitter, often older women who play a disproportionate role in spreading false information. For Huang et al. (2024), large language models can be seen as a factorization of the “culture” contained in the training base, where a “compressed” subject is a fact in the Wittgensteinian sense, although the truthfulness of facts is always debatable​.

Foreseeing, Predicting, or Imagining Worlds

Instead of rewriting the past, stories also allow us to imagine futures. For Peter von Stackelberg and Alex McDowell, there are four major classes of narratives in Science Fiction: “cautionary tales,” “thought experiments,” “literalized metaphors,” and “explorations of new science and technology.” The first corresponds to warning stories highlighting the—generally negative—consequences of some aspect of today’s society. The second consists of thought experiments that examine the potential impacts of a current or anticipated event, technology, or trend. The third uses literal metaphors to study a particular aspect of our world—for example, stories of aliens addressing dictatorship in our society—making it concrete and understandable. Finally, the fourth offers explorations of new sciences and technologies, using new advancements as the basis for a plot. These world-building stories are like maps that allow us to test our current strategies, discover new opportunities, and avoid threats. In The Ministry for the Future, Kim Stanley Robinson (2020) has scientists, policymakers, and activists debate political change. These imaginary exchanges are interesting and enlightening, but it’s mainly the power of the novel that makes us feel involved. We identify with the characters and live through the heatwave with them. After a (real) heatwave in the United States, James Nickell reported this sentence: “It’s easy not to care when you’re not the person sitting in ninety-five-degree heat.” Indeed, air-conditioned houses, cars, and offices allow many decision-makers never to feel the impact of a heatwave firsthand. The novel enables us to imagine what temperatures of 50 degrees (or more) truly mean.

In Silicon Valley, science fiction writer Alain Damasio delivers chronicles of his journey to Silicon Valley—a mix of a journal kept in 2024 by a French author and improbable encounters with characters and objects that seem to come from the future. Thomas Michaud pointed out as early as 2010 the strong entanglement between major IT companies and cyberculture artists’ laboratories: “The ideology of science fiction constitutes a unifying and structuring discourse within technoscientific capitalism,” with storytelling becoming a forward-looking vision capable of stimulating investors, employees, and consumers. This function of attracting investment is prevalent in the rhetoric accompanying tech startup fundraising efforts. “Unicorns” rarely communicate about rational and scientific aspects, preferring to highlight dreams and wonders (and probably also glitter).

A 2006 report on the ethical issues of nanosciences and nanotechnologies from the CNRS, in France, went even further, analyzing the link between science and science fiction: “Science fiction thus contributes to an ideological process influencing scientific evolution. It generates a shift in logical systems, raising questions about scientific normality and the possibility for engineers and scientists to explore fictional universes where traditionally accepted scientific laws are subverted and challenged. Finally, science fiction fosters thought experiments for engineers, who then view the known universe as a place likely to hold hidden truths that must be uncovered through science. Individuals in a society believe they act according to their free will, but are often included in an ideology that determines their activities and coordinates societal logic. That is why it is useful to consider science fiction as an ideology structuring innovation, conceived as a collective phenomenon governed by technocrats who succeed one another and accompany periods of innovation, often resulting from an important scientific or technical discovery.”

Thus, the story is not just there to sell dreams to potential investors. This fiction serves multiple purposes. In an uncertain world, it fulfills an epistemological function similar to that of thought experiments. It raises awareness of problems, playing a regulatory role. It also has a heuristic function, sometimes explaining the sequence of causes and effects. Finally, it serves a social function, enabling (or initiating) a dialogue between researchers and the public.

Beyond Numbers: Imagining Counterfactual Situations

In The Book of Why, Judea Pearl and Dana Mackenzie (2018) advocate for a new scientific approach that fully integrates causality, allowing for better understanding and decision-making in various fields. This is the idea behind most policy evaluations, to determine whether a decision made was good (or ideally, the best). For example, we might wonder if locking down a country for several weeks was the right decision. Or, we can ask what the health or economic consequences might have been had we not locked down in spring 2020. Looking from a slightly different perspective, David Hume argued in his Treatise of Human Nature (1739) that one cannot deduce normative statements (“what should be” – “ought”) from simple descriptive statements (“what is” – “is”), historically observed norms, or statistical averages—what the data would say if they could speak. Constructing counterfactual scenarios is not meant to answer a normative question (“what should be” for moral or ethical reasons) but to understand the potential consequences of our actions (“what should be” through logical reasoning). Particularly in policy-making, “evidence-based” policy is the gold standard. When we address an ultimate decision-making authority, we want to specify the lives saved, the budgetary impacts compared to public savings, and how current investments prevent costly problems in the future. Evidence is supposed to rely on a model of causality, and complex (or wicked) problems, by definition, have weak causal models. Stories, therefore, serve two purposes: they allow us to reason where data and theory fail. On the other hand, policy depends on the public’s will, and stories offer a rallying point, unlike facts and figures.

Comparing reality with the counterfactual allows us to understand the consequences of an action. This is the basic reasoning when we want to implement risk prevention strategies. As Eva Illouz points out, emotions shape our perception of reality and influence our decisions. Imagining counterfactual scenarios also means mobilizing our emotional capacity to feel and understand the implications of our choices. The TV series La Fièvre, analyzed by Raphaël Llorca and others (2024), illustrates how fictional situations can shed light on real issues by eliciting strong emotions that prompt reflection and action.

Constructing these counterfactuals often relies on very strong assumptions, and without them, causal reasoning is impossible. We are then very close to thought experiments or “Gedankenexperimente” as defined by Ernst Mach at the end of the 19th century. Alexandre Koyré and Karl Popper recall the importance of these thought experiments, even for sciences as serious and experimental as physics. Aristotle taught that larger stones fall faster than smaller ones simply because that is what the facts, experiments, and data said. It wasn’t until Galileo proposed an imaginary experiment that this notion was challenged. For Karl Popper in The Logic of Scientific Discovery (1959), this is linked to the fact that between empirical data and the theoretical object, there will always be an unbridgeable gap. In the case of falling stones, it is the difficulty of conducting experiments without air resistance. And it is precisely at this point that imagination and storytelling come into play by proposing to eliminate this gap, allowing us to imagine an ideal world, manipulating theoretically perfect objects, and serving as intermediaries between the mathematical and the real. In other words, for Alexandre Koyré, Aristotelian physics is not mathematical physics, and thus should not be taken literally. Scientific thought cannot occur without imaginary experiments.

Provoking Emotion to Drive Action

Ulrich Beck in Risk Society (1986) asserted that “science is becoming increasingly necessary but less and less sufficient to develop a socially established definition of truth.” In other words, science alone cannot suffice for making collective decisions. Scientific controversies become public issues, and political decisions rely increasingly on scientific expertise while also going beyond it. What’s more, scientific uncertainty often causes fear, especially when we begin to quantify that uncertainty. Who hasn’t been frustrated after a beautiful sunny day when an outing was canceled because rain “with a 90% probability” was forecasted? “Someone told me that each equation I included in the book would halve the sales,” said Stephen Hawking.

In the early 1970s, Herbert Simon defined the concept of the “attention economy,” highlighting the value of attention as a scarce resource in a world saturated with content. This theory can help us understand how to communicate scientific issues, where complex and often abstract concepts must compete with a plethora of appealing but less critical content. Emotions play an essential role in capturing attention and prompting action. At first glance, messages that evoke fear or a sense of urgency can attract immediate attention and action. And that fear must be as concrete as possible. Announcing exponential growth in disasters does not scare people because the announcement remains too abstract. Stimulating emotion is not just a communication strategy; it is also a necessity in an information-overloaded landscape, as road safety campaigns regularly remind us. Ronald Rogers shows that as early as the 1960s and 1970s, theories were developed that discussed the potential behavior changes induced by punishment or fear. Koehler et al. (2022) show that if we simply instill fear, these strategies are counterproductive. It is essential to suggest courses of action. Delphine Ernotte Cunci, in her analysis of the quality of information during the “fever,” highlights the importance of investigation and in-depth analysis to counterbalance the culture of immediacy. By integrating emotion and imagination into our decision-making process alongside numbers, we can transform these reflections into concrete and effective actions.

1 The exact citation is: “When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind.” A shorter version of this was engraved on the facade of one of the buildings at the University of Chicago. During a symposium in 1939, the economist Frank Knight humorously suggested replacing the quote with “If you cannot measure, measure anyhow.”

2 Called “במדבר” (Bamidbar, meaning in the desert) in Hebrew, “Ἀριθμοί” (Arithmoi, meaning numbers, which later gave rise to the French word “arithmétique”) in Greek, or The Book of Numbers in English.

References

Arditi, Andy, Balcells Obeso, Oscar, Aaquib111, Gurnee, Wes & Nanda, Neel (2024). Refusal in LLMs is mediated by a single direction. AI Alignement forum.
Barthes, Roland (1966). Introduction à l’analyse structurale des récits. Communications, 8(1), 1-27.
Brill, Steven (2024). You Think You Know How Misinformation Spreads?, Wired, 30 May 2024.
CNRS, Comité éthique (2006). Avis enjeux éthiques des nanosciences et nanotechnologies, rendu le 12/10/2006.
Ernotte Cunci, Delphine (2024). Soigner l’information au temps de la « fièvre ». Commentaire n° 186, 313-319.
Hackenburg, K., Tappin, B.M., Röttger, P., Hale, S., Bright J., Margetts, H. (2024). Evidence of a log scaling law for political persuasion with large language models. arXiv:2406.14508.
Hadjadj, Fabrice (2020). L’épidémie des chiffres, David et Covid. Vidéo Youtube.
Huang, Yuzhen, Zhang, J., Shan, Z., He, J. (2024). Compression Represents Intelligence Linearly. 15 April 2024. arxiv.org/pdf/2404.09937
Illouz, Eva (2022). Les émotions contre la démocratie, Premier Parallèle.
Koehler, M., Brockamp, T., Bamberg, S., & Gehlert, T. (2022). Change of risk behaviour in young people–the effectiveness of the trauma prevention programme PARTY considering the effect of fear appeals and cognitive processes. BMC public health, 22(1), 595.
Koyré, Alexandre (1960). Le De Motu Gravium de Galilée. De l’expérience imaginaire et de son abus, Revue d’Histoire des Sciences et de leurs Applications, vol. 13, no 3,‎ 197–245
Lewis, I., Watson, B., Tay, R., & White, K. M. (2007). The role of fear appeals in improving driver safety. International Journal of Behavioral Consultation and Therapy, 3(2), 203.
Llorca, Raphaël et al. (2024). Sur la fièvre. Enseignements politiques d’une série. Fondation Jean Jaurès.
McCloskey, D. N. (1990). If you’re so smart: The narrative of economic expertise. University of Chicago Press.
Mayer-Schönberger, Viktor & Cukier Kenneth (2014). Big Data: A Revolution That Will Transform How We Live, Work, and Think. Houghton Mifflin Harcourt
Michaud, Thomas (2010). La stratégie comme discours : La science-fiction dans les centres de recherche et développement. L’Harmattan.
Nickell, J.K. (2024), The Eviction Cure. Texas Monthly, July 2024.
Rogers, R. W. (1975). A protection motivation theory of fear appeals and attitude change. The journal of psychology, 91(1), 93-114.
Simon, Herbert. A. (1969). Designing organizations for an information-rich world. Brookings Institute Lecture.
Smith, Lynn (2001). Not the same old story. Los Angeles Times, 11 novembre 2001.
Timmer, John (2024). Key misinformation “superspreaders” on Twitter: Older women. Ars Tecnica, 30 May 2024.
Von Stackelberg, Peter, & McDowell, Alex (2015). What in the world? Storyworlds, science fiction, and futures studies. Journal of Futures Studies, 20(2), 25-46.

---

---