High-stakes decisions do not require narrative conviction but narrative flexibility

Abstract

We challenge Johnson et al.'s assumption that people reduce unclear situations to a single narrative explanation and that such reduction would be adaptive for decision-making under radical uncertainty. Instead, we argue that people imagine and maintain multiple narrative possibilities throughout the decision-making process and that this process provides cognitive flexibility and adaptive benefits within the proposed model.

The target article by Johnson et al. provides a framework for decision-making under radical uncertainty that relies on narratives to create the “single likeliest” chain of events (target article, sect. 3, para. 11). However, it does not leverage a central feature of narrative thinking: Our ability to imagine multiple possible pasts and futures. While Conviction Narrative Theory (CNT) posits the simulation of more than one imagined future, it also assumes that explanatory reasoning results in a single narrative that is “retained for downstream computations such as forecasting” (target article, sect. 7.2, para. 2) and that simulation generates “only one future for each choice” (target article, sect. 3, para. 7). We challenge the assumption that people reduce unclear situations to a single narrative explanation and that such reduction would be adaptive for decision-making. In other words, we disagree that “stories resist Heisenberg's principle – they take only one state at a time” (target article, sect. 7.2, para. 5). We contend that the plurality of narrative thinking should be integrated more fully into the various processes of CNT: Not only do we simulate more than one imagined future for each choice, but we also make sense of the present by relying on multiple narratives at the explanatory stage of decision-making.

Let us consider the authors' example of the widow hearing a noise inside her home, with a slight modification, to illustrate the importance of multiple narratives in decision-making. The widow hears a noise coming from downstairs late at night and considers two competing explanations: The noise could be a burglar breaking in, or it could be her lover entering unannounced, but she cannot know for certain. Even if she adopts whichever narrative “feels right,” the alternative explanation remains in the back of her mind. She then simulates several possible courses of action: She wants to defend herself; however, if it is her lover, she does not want to attack them. If the widow follows CNT, she would make a choice and bet on “the single likeliest imagined future” (target article, sect. 3, para. 11). Either she would grab and swing her baseball bat or she would jump forward to hug her lover. Obviously, a false choice here would be dangerous.

This example highlights that in many situations, it may be necessary to simultaneously consider not only multiple simulations of the future, but also multiple explanations of the present. If narratives play a key role in decision-making, then to understand this process we should consider how people make sense of them. How people consume narratives does not necessarily indicate how people produce them to make decisions. However, the process of narrative consumption provides insight into the relevance of narratives for decision-making because receiving a narrative also involves interpreting an ongoing situation and making predictions for the narrative's future (Campion, 2004; Magliano, Dijkstra, & Zwaan, 1996).

In research on narrative processing, there is in fact evidence that people do consider multiple possible versions of narratives simultaneously. Every narrative involves expectation management for the future, and thereby includes at least a minimal degree of suspense. Suspense involves simultaneously and continuously imagining at least two possibilities for a narrative's future, namely a feared outcome and a desired outcome (Carroll, 1990). Intriguingly, suspense and narrative enjoyment remain largely intact even when the outcome of a narrative is known (Carroll, 2001; Johnson & Rosenbaum, 2015; Leavitt & Christenfeld, 2011). One potential explanation for this paradox is that the narrative consumption experience is not only about accurately predicting a single outcome of the story, but also imagining alternative possibilities (Hiskes, Hicks, Evola, Kincaid, & Breithaupt, 2022). This view of narrative processing as possibility generation helps explain the popularity of fan fictions that aim at playing-out alternative versions of a story, as well as economic scenarios involving rapidly shifting opinions about Bitcoin, to use Robert J. Shiller's example (2019), since competing narratives are already available.

Just as narrative processing research indicates that people may hold multiple future narratives in mind simultaneously, there is also evidence that people maintain possible alternatives for a narrative's past using counterfactuals. De Vega, Urrutia, and Rifo (2007) find that immediately after reading, counterfactual and factual information remains separate but accessible in memory to readers. Similarly, Ferguson and Jayes (2018) find that readers do indeed attend to and evaluate counterfactual information, even when it differs from factual narrative situations. Consequently, there is evidence that when consuming narratives, we do not remain “blind to alternative possibilities” as Johnson et al. contend (target article, sect. 7.2, para. 1). Instead, alternative explanations of a situation remain open and available.

While this research addresses actual narrative processing, we argue that considering multiple possibilities in real-life decisions is also adaptively beneficial. The authors insist on the value of conviction in decision-making, while highlighting the subsequent need for learning and reassessment based on new information. However, conviction may be detrimental in quickly changing situations, and constant reappraisal and reassessment may be costly as well. If a situation is radically uncertain, settling on a single narrative explanation may pose a high risk, as the example with the widow demonstrates. Instead, keeping multiple possible narratives in mind minimizes the chances of being surprised by, and thus unprepared for, changing circumstances. Moreover, such narrative multiplicity offers a way to compare and discriminate between different courses of action at both the evaluation and simulation stage. In other words, keeping multiple narratives open in decision-making lowers the risks posed by overcommitting to a single narrative that may ultimately offer bad courses of action.

Thus, we recommend a modification to CNT: Considering multiple narrative possibilities throughout the decision-making process is not only something that people do, but it is also beneficial for evaluating the best possible decision. This multiplicity does not preclude decision-making but allows for flexibility as we try to understand our present through explanation, imagine our future through simulation, as well as make decisions through affective evaluation.