Lingers
Rocco Mickel
Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Lingering should be elicited both by the properties of the stimulus and by our appraisal of it24,31. Stories contain high-level semantics and situational information, and so the process of comprehending a narrative is likely to require deep meaning-centered processing, leading to lingering. However, deep processing can also arise from endogenous interests and dispositions toward the world: we may struggle to interest ourselves in an episode of a popular TV show, while at the same time finding ourselves engrossed in the plight of an ant struggling to carry an outsized breadcrumb down from our armchair. We predict that the extent to which we engage deeply in our thinking about the ant, specifically considering it in terms of the broader situation or narrativizing its activities (e.g., where is it going and why is it so motivated?), will cause ant-related thoughts to linger after the ant leaves our immediate perception.
When participants read coherent narratives, the themes and details of the story lingered for several minutes in their subsequent free association chains, more so than in participants who read scrambled versions of the same text. This observation replicated and generalized across multiple stories. A follow-up experiment demonstrated that coherent stories lingered more when participants judge the emotional properties of the sentences rather than their spelling or font, and also that the lingering was mostly nonvolitional. Moreover, the lingering effect was again observed when participants narrativized a word list, but not when they judged the perceptual properties of each individual word (e.g., italic type). Overall, regardless of the objective coherence of what participants read, their subsequent experience of lingering was best predicted by the degree to which they felt transported by the material while reading.
What determines whether a past experience persists in our mind? Despite the prominence of history-dependence in our models of the human mind and memory5,6,8,38,39, we have little empirical evidence to support our intuitive sense that meaningful experiences resonate with us, shaping our thoughts in the minutes after they end. Here, we empirically demonstrate that, when people interpret a text as a narrative of interconnected situations, instead of focusing on the semantic or perceptual features of individual sentences or words, they experience a lingering influence on the trajectory of their subsequent thoughts for minutes afterward.
Why should attending to situation-level meaning elicit a lasting mental context? A potential explanation comes from the levels of the processing framework of human memory26,53. The levels of processing framework stems from work on perception, where the perceptual process was conceptualized as a hierarchical series of tests at different levels of analysis54. Early levels are concerned with the physical properties of a stimulus, while later levels examine more abstract stimulus properties such as meaning and implication. According to levels of processing, the persistence of the stimulus in memory is a function of these levels of analysis: stimuli that are processed at later (deeper) levels are more likely to form lasting representations in memory. For example, studying a word list by deciding whether or not each word is capitalized results in poorer recognition than studying them based on their fit in a sentence55. However, the depth of processing does not stop at word-level semantics. We contend that engaging with the situation-level meaning of a text is a prime example of even deeper meaning-centered processing36 and should result in persistent representations in memory. While levels of processing models traditionally concern persistence in memory rather than spontaneous thought, there is reason to consider these constructs are related. For example, overall memory performance is positively correlated with history-dependence (i.e., temporal clustering) in freely recalled word lists56. Also, the extent to which a recent social experience permeated thoughts during a post-task rest period predicts subsequent memory for the original experience19. Lingering in spontaneous thought may be a natural consequence (or antecedent) of robustly encoded memories.
Precisely how deep processing results in lingering remains an open question53. If we consider all of the individual units of our knowledge and experience as nodes on a graph, deep processing may be operationalized as a learning function that results in lasting increases in the edge weights between the nodes of an input and the nodes of related knowledge and experiences. If we then model spontaneous thought as a random walk on such a network57, we should be more likely to traverse these deeply processed edges again (i.e., lingering). Furthermore, the consequences of deep processing according to this model would not be limited to non-volitional random walks, but would also benefit more rule-based search processes on the network, consistent with the well-documented benefits of deep processing on tests of explicit memory26. However, how a pure associative network model could support the lingering of more complex mental representations, such as our current concerns or goals58 remains unclear. Agent-centered models59,60 that combine decisional and episodic memory processes may be necessary to capture the real-world phenomenon in which our thoughts during one task or interaction persist for minutes into the next task or interaction. In the brain, deep processing may result in lingering via a propensity to drive activity in higher-order association cortices (e.g., regions of the DMN)38,39. Higher-order association cortices possess distinctively slow-drifting intrinsic dynamics, likely due to their elevated levels of local-circuit and inter-regional recurrence61,62. Thus, if deep processing especially involves these brain regions, they are well-placed to generate lasting neural reverberations and, perhaps, lingering mental contexts.
Philosophers and psychologists have noted that our stream of thought echoes recent and distant memories, and that each moment informs the meaning of the next1,2. Here we demonstrated that the extent of this history dependence is not a fixed parameter. Instead, the extent to which our recent past lingers into subsequent thought increases as a function of processing depth26,53. The more we consider the deep situation-level meaning of an experience, the more likely it will exert a lasting mental context and shape the trajectory of our subsequent thoughts.
Participants progressed through the text at their own pace by pressing space bar after reading each sentence. All stories were between 2,158 to 2,798 words in length, ranging between 196 and 268 sentences.
To determine whether classification accuracy was above chance, we generated a null distribution of 500 accuracy values. The null accuracy values were generated using the same procedure as described above, after randomly shuffling the pre- post- labels from the test dataset for each fold in cross-validation procedure. We then computed the proportion of null accuracy values greater than the empirical classification accuracy.
Procedures were identical to Experiment 1, except that the self-paced reading phase was replaced by incidental list-learning and Experiment 2 included additional post-story components: story description and a test of free recall.
Next, participants completed an edited version of the Narrative Transportation Questionnaire35, which sought to measure the extent to which they were transported into the hidden story. Then, participants were asked to (i) type a summary of the hidden story in their own words, and then performed (ii) a free recall test, in which they were asked to recall as many of the words from the original word list as possible. During free recall, participants typed words into the center of a blank screen and pressed Enter to submit them. After pressing Enter, the word disappeared. After the free recall, participants the identical multiple choice comprehension test used in Experiment 1 for the Carver story. Finally, participants answered questions about their demographics, the strategies they used, and their subjective experience of lingering, all using an identical format to Experiment 1.
769 participants took part in Experiment 2 and were recruited via Amazon Mechanical Turk. Data were collected during July 2020. Participants were paid $6.00 USD for their participation and provided informed consent before participating.
Experiment 4 was a preregistered follow-up experiment ( ; January 21st, 2022) that sought to determine whether we could limit how deeply we process a coherent narrative, and thereby reducing lingering? Procedures were largely identical to those in Experiment 1, except that the self-paced reading task was modified such that each sentence was presented in the context of one of two cover tasks, manipulating the depth with which participants read an intact story. All participants read a version of the intact Carver story, where half were randomly assigned to a condition that encouraged shallow processing of the story (i.e., proofreading the text for spelling and font errors) and another that encouraged deep processing (i.e., rating the valence of each sentence in the story). For more details, see SI: Supplemental Methods, Supplementary Note IX and X. All procedures followed those in the preregistration protocol, with the exception of an additional participant exclusion criterion ensuring that participants in the Proofread condition were indeed detecting errors above chance. This criterion was accidentally omitted from the original protocol.
7fc3f7cf58