The wisdom in the story: Clarifying assumptions about radical uncertainty and reasonableness in narrative judgment

Abstract

Human lives are radically uncertain. Making sense of such uncertainties is the hallmark of wisdom. Sense-making requires narratives, putting them in the center stage of human everyday decision-making. Yet what if radical uncertainty is a narrative itself? Moreover, do laypeople always consider such narratives irrational? Here we pose these questions to enrich a theory of choice under uncertainty.

Everyday human decision-making is neither purely analytic and probability-driven nor is it about whimsical constraint satisfaction. There is nothing radical about this claim – most of the world's philosophical traditions put context-sensitive meaning-making instead of logical deduction of correct probabilities at the heart of discourse about what makes up human wisdom (Grossmann, 2017). After all, the ecology of everyday life is rarely well-defined: Preferences of other decision-makers are rarely transparent, and situations typically evolve. Consequently, logically derived probabilities may be of little use for decision-making in the ill-defined contexts of everyday life. Instead, at least since Aristotle scholars have emphasized meta-cognitive faculties that help one to navigate life's uncertainties, get a better sense of the issue at hand (Grossmann, 2017; Grossmann et al., 2020b), and to justify their choices (Grossmann, Eibach, Koyama, & Sahi, 2020a).

In the Conviction Narrative Theory (CNT), Johnson et al. echo these philosophical themes, leveraging insights from cognitive, affective, and behavioral sciences to shed light on how narratives guide human choices under “radical” uncertainty. They postulate a powerful and insightful account of narrative reasoning and the conditions under which people are likely to use it. We applaud the authors' ambitious attempt to integrate disparate bodies of literature in decision sciences and philosophy. At the same time, for CNT to become a common vocabulary in decision sciences, it may need greater clarity about two of its foundational components: Radical uncertainty and reasonableness (versus rationality) as a criterion of good judgment.

CNT postulates that people resort to narrative reasoning in situations where they face radical uncertainty, in which the probabilities of outcomes (or evaluative criteria) are unknowable or ambiguous. Below, we outline how radical uncertainty and fuzzy evaluation may themselves sometimes be the product of narrative reasoning. The person's causal models of the world can determine whether they perceive a given decision problem as a situation of manageable uncertainty, where reasonable estimates of probabilities and values can be assigned to decision outcomes, or a situation where the probabilities of outcomes are too radically uncertain or fuzzy to evaluate using probabilistic reasoning.

Causal models of the world can determine how much uncertainty is perceived in any given decision problem because such models determine what range of factors need to be accounted for. Some folk theories may lead people to situate a particular problem within a relatively tractable deterministic, linear casual system where uncertainty is calculable. For example, Enlightenment thought conceived of a world whose properties were rationally ordered and, at least in principle, knowable. By contrast, other folk theories may situate the same problem within a dynamic, non-linear system in which radical uncertainty applies. For example, postmodern narratives conceive of a world in which most or all knowledge claims are fundamentally suspect and subject to shifting power dynamics within society. Thus, radical uncertainty is the exception within the Enlightenment worldview but the norm within the postmodern worldview.

Notably, there are individual and cultural differences in people's acknowledgement of the uncertainty and changeability of events and awareness of limitations in their knowledge and understanding, which are the core epistemic premises of wise reasoning (e.g., Grossmann et al., 2012; Porter et al., 2022; Santos, Huynh, & Grossmann, 2017). Such differences in lay epistemic assumptions will mean that some decision-makers construe a particular judgment as a situation of radical uncertainty that requires narrative reasoning approaches while others construe it as a situation of manageable uncertainty in which probabilistic reasoning is applicable. Moral narratives may also frame probabilistic reasoning as taboo in certain contexts, such as making decisions involving sacred values (Baron & Spranca, 1997; Tetlock, Kristel, Elson, Green, & Lerner, 2000). Thus, even probabilistic reasoning is ultimately contingent on prior narratives – namely, narratives that frame the judgment as one in which there was relevant, reliable, sufficient information to assess and evaluate the probabilities of outcomes (Douglas, 2001).

Moreover, individuals may be tactically motived to construe a situation as if involving radical uncertainty. In a recent example, even after there was compelling clinical evidence of the efficacy of COVID mRNA vaccines for reducing COVID illness, some vaccine skeptics framed vaccination as a context of radical uncertainty by emphasizing that until long-term data on vaccination outcomes were available it was impossible to know whether vaccination is in one's interest (Lu, 2022). Some people expressed hesitancy about getting a COVID mRNA vaccine due to a variety of imagined long-term side effects ranging from cancer to infertility (Pertwee, Simas, & Larson, 2022). Radical uncertainty can also be rooted in conspiracy narratives that discount official information about probable outcomes as fabrications designed to manipulate the population for nefarious purposes, such as the conspiracy theory that public health information about vaccination efficacy was an elaborate hoax to get tracking microchips into people's bodies (Pertwee et al., 2022). These examples illustrate how radical uncertainty can be a subjective product of narrative reasoning rather than an objective description of the available information.

Our second point concerns criteria of good judgment within CNT. CNT is presented as a type of bounded rationality where the benchmark for sound judgment is reasonableness rather than rationality. Hereby, the definition of reasonableness appears to rely on folk concepts (Grossmann et al., 2020a), whereas the definition of rationality is a smorgasbord of theoretical positions not derived from laypeople. Johnson et al. argue that most definitions of rationality rely on assigning probabilities, which are absent under radical uncertainty and thus disqualify rationality as a possible benchmark.

This argument assumed that the definitions of theorists and of laypeople are equivalent, which is not necessarily so. Lay people define rationality as reductionist and seeking to serve a single value whereas they define reasonableness as holistic and seeking to balance incommensurable values (Meyers, Eibach, Hanxiao, & Grossmann, 2022). The lay definition of rationality does not require probabilities to be assigned, sidestepping the original rejection of rationality as a benchmark. If the lay standard of reasonableness is used for CNT, then the lay standard of rationality should be used, too. This would parallel non-radical decision scenarios wherein people hold both standards to be descriptions of good judgment, dependent on the goal(s) of the decision-maker.