Flow and longform reading. Notes from an exploratory case study

1 Introduction

Recently, both international and country-specific survey studies indicate that students increasingly struggle with longform reading (Baron and Mangen 2021; Hakemulder and Mangen 2024; Mullis et al. 2023; OECD 2023). Many seem to find it difficult to engage with lengthy and complex texts, such as those required in academic courses. Commonly labeled “deep reading” (Wolf and Barzillai 2009) or “higher-level reading” (Schüller-Zwierlein et al. 2022), such reading goes beyond the simple task of decoding texts and retrieving explicitly stated information. It encompasses processes such as complex inference-making based on often implicit information, analogical reasoning, analytical and critical reflection, evaluation of opposing points of view, exercising sustained attention and what Maryanne Wolf has termed cognitive patience (2018) – a type of cognitive endurance and grit. Deep or higher-level reading skills are a cornerstone of many academic programs, but they are equally crucial for engaging with other types of texts, such as literary works, essays, political documents, biographies, policy papers, and historical or legal texts. More broadly, in an era of constant audiovisual stimuli and information overload, deep, longform reading serves as a powerful way to cultivate sustained focus and concentration (Schüller-Zwierlein et al. 2022). It also acts as a gateway to critical thinking, analytical reasoning, and perspective-taking (Wolf 2018).

Hence, if the decline in deep reading habits and skills is sustained in coming years, a wealth of cultural, academic, and other forms of knowledge that is inscribed in lengthy and complex texts will be lost to a growing number of people. Finding ways to facilitate and scaffold deep reading is therefore a key challenge in today’s information society. A prerequisite is to map the conditions – in the environment, of the reader, and of texts – that are conducive to deep reading. One of these is flow, an experience of being fully engaged in an activity, theorized to depend on an optimal balance between individual challenge and individual skill levels, and associated with intrinsic motivation (Thissen et al. 2018). Deep reading requires the ability of the reader to “stay put” with textual information of various degrees and types of complexity, over an extended period of time. When successful, such sustained focus with a text may bring about a sense of deep and immersed engagement where elements of the surroundings recede to the background, time is entirely experiential, and the reading is automatic and more effortless than during moments when the reader is struggling, for instance, with comprehending the text, keeping distractions in the environment at bay, or countering boredom or restlessness.

With these stakes in mind, we wanted to explore the conditions that facilitate and inhibit flow during longform reading among university students, and, on the basis of such exploratory case study, design a pedagogical program to help students improve their ability to engage in longform academic reading. Our study was in part motivated by the wish to explore the impact of having access to smartphones and other internet-connected devices during reading. Given that people constantly carry their smartphone with them, the mere option of checking an incoming text message may be seductive (cf. Gazzaley and Rosen 2016) and disturb the possibility for maintaining focus on reading. Switching between textbook reading and engaging with an app forces readers to constantly reorient their attention. This oscillation takes time and requires effort. This phenomenon, known as attention residue (Leroy 2009), where the mind remains in part preoccupied with the prior task, results in a delay of at least a minute or five. Attention residue is particularly prevalent when individuals switch between tasks without completing the initial one or when the tasks are complex and require significant cognitive resources. However, it does not occur with every mental activity; its likelihood increases with the complexity and emotional investment in the tasks involved. In addition, research has indicated that people involved in such task-switching take longer to complete their tasks than those who stayed focused on only the primary task (Rogers and Monsell 1995; Rubinstein et al. 2001).

Such challenges have only grown in recent decades, as the number of stimuli to which most people are exposed has increased exponentially, while the so-called attention economy is becoming more competitive (Bruineberg and Fabry 2022; Nelson-Field 2020). As part of our exploratory study, we therefore ask: What can we learn from observing how readers enact the optimal conditions for deep reading, involving immersion and flow, in natural settings, where readers read on an ordinary basis? And how do readers describe their subjective experiences of entering and sustaining flow-states while reading? By addressing these questions, we aim to uncover the interplay between observed behaviors and self-reported experiences, ultimately informing the design of practices that scaffold deep reading and foster optimal reading conditions. In the sections that follow, we (2) present the theoretical framework that informed our study, (3) outline the details of our methodology, (4) discuss our results and (5) limitations, and finally (6) we reflect on their implications for the scholarly literature, policy debates and future studies.

2 Theoretical framework: embodied cognition and flow in reading

In recent years, a new theoretical paradigm has emerged in reading research, converging on the understanding that reading is shaped by processes beyond the relation between a pair of eyes and a text (see e.g. Hillesund et al. 2022; Mc Laughlin 2015; Trasmundi et al. 2021). The distributed and embodied approach to reading suggests that reading is better understood as a dynamic task in which the whole body engages dynamically with the surrounding environment (Trasmundi 2024). Such a perspective reshapes the unit of analysis by taking into account how the experience of reading is defined by the wider context in which the activity occurs, including other people, the medium with which one reads, and the surrounding environment. When traditional cognitivist theories study flow processes in reading, the emphasis primarily falls on the psychological mechanisms that enable or constrain mental states conducive to reading (e.g. Nell 1988; Thissen et al. 2021). In this view, there is little need to consider the role and function of the body and its wider environment, because ‘flow’ is mainly considered a neurocognitive phenomenon. However, an embodied and distributed account of reading extends the unit of analysis to the brain-body-environment system (Hutchins 1995). Understanding reading as an embodied, contextual activity, requires investigations of relations and interaction between components in this wider system – in this case, how readers engage with texts and the broader reading ecology, encompassing prior experiences and expectations regarding the management of reading tasks, as well as reading aids, communication tools, food, and interactions with other people. The results, as we describe below, reveal the importance of studying the brain-body-environment system to better grasp how people get into the flow of reading, and what it may look like when they experience flow in reading.

But what, exactly, is ‘flow’? The concept and theory of flow were originally applied to experiences involving specialized physical expertise and engagement with the immediate environment, whether through artifacts and technological devices, as people would become completely absorbed in activities like chess, computer games, painting, or playing musical instruments, or through the texture and materiality of the natural world, as in rock climbing or swimming (Beard 2015). However, flow has also been applied to activities with seemingly limited physical enactment, such as reading. McQuillan and Conde (1996) combined interviews with the administration of the established Flow Questionnaire (Csikszentmihalyi and Csikszentmihalyi 1988) to identify the conditions under which readers report being intensely engaged in a text, losing a sense of time and place. They found that texts most commonly prompting a sense of flow in readers were those read for pleasure, with fiction texts more likely to induce flow than non-fiction. More recently, Thissen et al. (2018) adapted the widely used Flow Short Scale (FSS; Rheinberg et al. 2003) to the context of fiction reading, developing the first reading-specific flow scale, the Reading Flow Short Scale (RFSS). Thissen et al. used the RFSS to measure what they termed “optimal reading experiences” of fiction texts. The RFSS is a significant contribution to research on flow in reading. However, because it is a self-reported measure administered after the event, it is subject to well-known biases related to social desirability and memory distortion. It is also important to note an inherent limitation of self-report measures administered after the experience: since the very phenomenon under investigation is likely to be disrupted by investigating during the reading, it is necessarily always in hindsight that participants evaluate whether items like “I was completely immersed in what I was reading” (Thissen et al. 2018: 3) applied to their experience, and are only enabled to make their decision for their reading as a whole.

Moreover, reading takes substantial time, especially longform reading, which is the focus of the present study. It unfolds not as a smooth, linear process but often as a ‘messy’ activity of various embodied fluctuations and oscillations, such as pointing, moving, and regulating the reading pace (Trasmundi et al. 2022). During this process, experiential aspects such as absorption and flow will likely vary in response to shifts in the environment, whether from the text itself, the reader’s own embodied experience (e.g., hunger, thirst, restlessness, or sleepiness), or external factors such as the presence of a mobile phone or other people’s presence. Verbal self-report measures, whether prompted by researchers during reading (e.g., think-aloud protocols) or after reading (e.g., diary approaches), rely on the reader’s ability to articulate cognitive and affective-emotional experiences. Many of these experiences, however, may be elusive to consciousness, such as subtle affective states. Additionally, verbal prompts administered during reading inevitably interrupt the very process they aim to measure.

The RFSS’s focus on the text, the reader, and their interrelation also overlooks a central focus in our pilot case study, namely the role of the reading environment beyond the text-reader nexus. These elements may play a crucial role in facilitating flow during reading, whether for pleasure or study. Supplementing this text-reader focus, our study highlights how other factors in the reading environment, such as the reading device, other readers in the room, artifacts, and the reader’s embodied dispositions, also need to be taken into consideration. Addressing these factors requires redefining reading to include the reader’s embodied interaction with artifacts and people, and recognizing how cognition is distributed across texts, readers, and their environments. Our exploratory study seeks to capture the role of these additional, often overlooked, elements in the reading environment for the reader’s experience of flow during extended study reading sessions. To do so, we employ a type of data largely untapped in the study of reading flow: video observations of real-time reading combined with focus group interviews conducted after the reading sessions.

3 Methodology

The study’s participants (N = 13) were recruited from the MA program in Social Anthropology at the University of Oslo. The sample consisted of students aged 24 to 30. All participants were non-native English speakers reading in English; however, they were enrolled in an MA program with English as the sole language of instruction. Upon signing up, participants committed to two five-hour reading sessions at the University Library. This design and duration was based on our decision to mirror an existing reading seminar at the university, called “the Deep Reading Seminar”, which had been offered to undergraduate social anthropology students since 2019 (Wig and Naguib 2021). We set up our study as an exploratory case study to investigate reading behavior within this existing and naturally occurring reading setting.

The students were informed that the purpose of the study was to understand their experience with longform reading. Based on insights into the impact of digital devices on longform study reading (Baron and Mangen 2021) and the differences between screen and print reading (Delgado et al. 2018; Clinton 2019), we further designed our study to allow an in-depth exploration of the role of these devices in particular. Hence, on day 1, participants were allowed to use any device or technology they preferred during the reading session, with no restrictions on access to digital technologies. That is, they could use this technology both as a reading and/or entertainment device, whereas on day 2, they did not have access to any digital technologies, neither in the room nor during breaks. On this day 2, they were instructed to read the materials on print and leave phones and other devices turned off and out of reach.

Beyond this technology constraint, the same setup was maintained for both sessions. Each desk was numbered, and the students sat at the same desk in both sessions. They read for 5 h and could take breaks outside the room at will. They were assigned in total six scientific articles from the syllabus in their subject of study, three for each session. Two extra academic articles from the syllabus were provided as extra reading materials for those who finished the assigned readings. The decision to have them read texts from the syllabus was primarily pragmatic, allowing us to schedule the reading sessions as part of the students’ own assumed reading plans and activities. We also assumed that this could be a way to motivate students to participate in the study.

We collected data from three sources. First, four video cameras recorded both reading sessions from multiple angles. Two randomly selected participants, one male and one female, were recorded with cameras recording up-close video, approximately one meter from each reader, providing full-body shots. This gave the basis for the detailed case studies discussed below (Section 4.4). Given that the study is a pilot, we considered it sufficient for the present purposes to have close-up video recordings of only two participants. Second, we collected basic demographic information in a survey, completed before participants started reading. Third, following each session, we conducted approximately 70-min focus group interviews with all the participants, following an interview guide with open questions focusing on the chronology of the reading experience, practices of notetaking, experiences during breaks, and returns to the text.

As outlined, it is methodologically challenging to determine whether someone experiences flow during reading. Below we present and discuss our observations in the video data in light of the concept of flow in reading, focusing in particular on the observed transition time from a seemingly constrained behavior to a smooth task performance. As the reader enters a state of flow, the intervals between bodily adjustments become longer. Video observations reveal that more time passes between each time a reader takes their eyes off the text, reaches for their phone, takes a sip of water, etc. As will become clear, however, we do not associate flow in reading with complete bodily stillness. Part of what happens as readers “settle in” and experience flow, is that they move “with” the reading devices at hand, picking the papers up while often sitting back in the chair with materials in hand, as shown in the video material described below. Because of the methodological challenges of studying flow as it occurs among readers, part of our contribution to the literature is developing a research procedure and vocabulary with which to describe and understand conditions and experiences of flow in longform reading to be used in future studies.

4 Results and discussion

Our study was designed to explore if and how reaching a state of ‘deep reading’ is associated with the presence or absence of digital mediums, as well as looking for observational differences when reading for a longer period on paper and digital devices. However, parts of our results stand out by demonstrating similarities across the digital and analogue settings. Below, we begin by describing commonalities between participants reading with and without access to digital tools, before moving to their observed differences. Finally, we turn to a detailed study of the two participants who were filmed up-close with video-cameras, allowing us to further tease out the details of reading in context.

5 Limitations

Before discussing the implications of these results, it is helpful to first acknowledge the methodological constraints of this study – including its small scale, and our reliance on rather few observations and self-reports. The nature of our data does not allow robust conclusions with respect to questions such as how long the readers actually spent “on task” during the overall 5 h of the dedicated session. Based on the video observations, we can show the frequency and the duration of breaks and pauses during reading, and whether these breaks and pauses seemed to be self-initiated or prompted by disturbances in the surrounding environment. Findings from our focus group interviews indicate that participants perceived their reading as more successful in the paper condition compared to the digital condition. However, since we did not measure their actual reading time or performance, our conclusions are limited to the participants’ self-reported experiences following the reading session, as well as our coding and analysis of the video material.

Such limitations show the importance and necessity of supplementing observational studies and self-report data with additional measures. To provide insights into what occurs during reading and allow for analysis of the effects of various types of interruptions or breaks on outcomes, such as comprehension. In order to shed more light on what goes on during reading, and to what extent readers are in fact reading or doing something unrelated, a more robust research design would supplement video data with physiological measures such as eye tracking, allowing more accurate identification of time “on task”. Recent advances in eye tracking equipment such as lightweight eye tracking glasses make it possible to track eye movements during reading with sufficient precision in natural settings. In addition, exploratory observational studies such as this current study could benefit from the inclusion of some performance-measure administered at the end of a reading session or an entire seminar, assessing students’ comprehension or recall of the assigned texts. In this way, we would be able to further specify the role of the reading devices, the quality of flow, and the embodied reading trajectories and how they contribute positively and negatively in study reading beyond the readers’ subjective experiences.

Another central question is whether or to what extent our findings relating to the fluctuations of flow are specific to reading, as compared to other tasks of the same duration, for instance, physical exercise, writing and drawing, playing various games and instruments, or relaxation practices such as yoga and meditation or even listening to storytelling. Granted, flow has been studied in many such activities (Høffding et al. 2024), and it would be interesting to compare reading-specific flow in comparison.

Lastly, it can be questioned how and to what extent our choice of text in the present study had a bearing on the results. Given that flow has been found to be more easily facilitated by fiction reading (McQuillan and Conde 1996), it is an interesting empirical question whether we would observe different patterns and reading trajectories if the participants were provided fictional texts instead of, as in our study, texts on their curriculum. Several textual parameters may be of interest in this respect. The fact that our participants read scholarly articles and not one long text, adds to the difficulty of comparing flow across the two reading sessions. Results from the focus group interviews also revealed that the preferences and perceptions of the students varied with respect to the different articles. Future studies exploring embodied reading trajectories and flow with paper versus digital texts should compare the reading of longer texts, with due scholarly attention to questions of genre and other textual parameters that may be important to flow (e.g., narrative or stylistic complexity; layout; information density; action-based or character-based stories).

6 Implications

While the recognition of the limitations of our study helps point in the direction of new questions, our results also suggest some possible answers that are relevant to reading researchers and policy makers alike. The results of our exploratory pilot study contribute to exploring our conceptual understanding of readers’ practice of concentrated engagement with longer texts. Based on our observations, we have discussed the importance of “settling in” as a prerequisite for achieving flow– an amalgamation of various factors related to the text, the reader, and the reading environment. We advocate considering a perspective that views reading performance as a process of bodily engagement with the environment, in which individuals adjust and modulate their bodily positions, aligning themselves with the task-environment and preparing the body as well as the mind to concentrate and engage. This process is partly facilitated by what the environment offers, be it the presence of other people engaged in long-form reading, or the accessibility of digital devices.

Encouraging or facilitating this process whether as individual readers or educators might enhance reading experiences. For instance, raising awareness of the phenomena discussed here, or providing opportunities for deep reading sessions in university libraries could improve the flow of reading, increase reading enjoyment, and ultimately lead to more pages being read within a given time. Such awareness could also facilitate access to longer and complex texts, both within and beyond academic circles, whether these are fiction or nonfiction. These implications align with findings from OECD studies mentioned in the introduction. They underscore the importance of deep reading skills in fostering comprehension, critical thinking, and the ability to engage with complex materials across academic and professional domains. Building on these results, policy makers, educational practitioners and scholars could consider how pedagogical practices and institutional settings might better support the development of these skills.

Beyond these potential implications for pedagogy and policy, our findings also lead to new research questions. Building on the insights from our pilot study, we suggest a broader research agenda aimed at systematically investigating the factors that positively enable deep reading and flow. This ambition could entail explorations of the interplay between physical and digital reading environments, examine the role of interruptions or breaks, and evaluate how these factors impact the subjective experience of flow and further measurable outcomes like comprehension and retention. By integrating these insights, future research could further inform embodied strategies to support deep reading in diverse contexts.

Two areas of future empirical research immediately strike us as particularly promising. First, building on the present pilot study, empirical studies could be carried out within educational settings to further investigate the effects of digital access and reading substrate on students’ flow. A sound research design would seek to control for confounds related to, for instance, differences in the content or type of the reading materials, the difference made by types of reading device (screen vs. print, screens with access to the internet vs. disconnected), as well as identifying relevant personality variables among participants, by establishing reader profiles to determine who benefits from digital device removal and who, potentially, does not. This empirical investigation could also seek to discern what constitutes the optimal length span of a deep reading session. Will readers adjust the timing of their manipulations and arrangements with artefacts, adjusting postures, hair etc. to settle into the length of the sessions?

Secondly, our study raises questions about the role of social facilitation in reading, and in particular, in challenging long-form reading. In focus group interviews, participants spoke enthusiastically about how social facilitation entailed by the group dynamic helped harness a sense of flow and task progress by providing a form of social support, an unspoken, mutual assistance. This topic merits further research. What effect does social facilitation have on reading flow? Does an individual reading experience lend itself to similar experiences of flow and focused reading as does reading as a collective experience?

In our view, collective deep reading holds potential not only as a subject of research but ultimately also as a form of pedagogy for students in higher education, if scaled correctly. Many students succumb to frustration or self-judgment when unable to read sufficiently, viewing these moments as their own failure to concentrate. However, the perspective we have developed in this article locates the “problem”, and hence also possible solutions, beyond the individual reader. An embodied and distributed cognitive perspective on reading reveals the significance of reading as a fully integrated, whole-bodied activity, where flow and immersion emerge when body, text, reading medium, and environment are harmoniously aligned. This alignment enables readers to engage deeply during reading, unlocking the transformative joy of getting lost in a text.