Preliminary steps towards a cognitive theory of fiction and its effects
- 77 Downloads
In the last few years there has been a real explosion of studies on fiction and its effects in social, cognitive, and media psychology, in communication science, in different subdisciplines of cognitive neuroscience, in experimental aesthetics, and in the numerically aided phenomenological approach to the study of fiction. This research is at an early stage and there is general consensus that it presents conspicuous shortcomings. In this paper I expressly address one of the most relevant limitations: the lack of an interdisciplinary integration among the different trends of research. I propose a theoretical-empirical review that integrates some of the most crucial findings concerning the fictional processing and its persuasive and learning effects across disciplines. The review is presented as a network of interconnected theoretical hypotheses, based on widely shared and well-researched conceptual constructs. Each hypothesis is supported by recent relevant findings and connects different lines of research. Taken together, the hypotheses represent a preliminary step towards a cognitive theory of fiction and its effect.
KeywordsSimulation Narrative Fiction Transportation Identification Epistemic vigilance Reflection
In the last few years there has been a real explosion of studies on fiction and its effects in social, cognitive, and media psychology, in communication science, in different subdisciplines of cognitive neuroscience, in experimental aesthetics, and in the numerically aided phenomenological approach to the study of fiction. This research is at an early stage and there is general consensus that it presents conspicuous shortcomings (Koopman and Hakemulder 2015; Oatley 2016; Panero et al. 2016; Barnes 2017; Burke et al. 2016; Oatley and Djikic 2017; Samur et al. 2017). In this paper I expressly address one of the most relevant limitations: the lack of an interdisciplinary integration among the different trends of research.
I propose a theoretical-empirical review that integrates some of the most crucial findings concerning the fictional processing and its persuasive and learning effects across disciplines. The review is presented as a network of interconnected theoretical hypotheses, based on widely shared and well-researched conceptual constructs. Each hypothesis is supported by recent relevant findings and connects different lines of research. Taken together, the hypotheses represent a preliminary step towards a cognitive theory of fiction and its effect.
On the basis of the available evidence, the review aims at identifying something akin to a prototype of the fictional processing and the multiple realizations of it. This prototype can be applied to very different social categories of fiction, stemming from the typical phenomena investigated by media psychology (in particular, I refer to evidence concerning advertising and educational entertainment) to artistic phenomena (in particular, I refer to literature, but I also present evidence concerning paintings, films, television shows, and video games). The prototype and its multiple realizations allow us to catch the persuasive and learning effects of fiction on very different cognitive processes: beliefs, emotions, mind reading, attitudes, and reflection.
In my view, the prototype of the fictional processing is based on four interconnected properties: simulation and believability, transportation and identification. These properties are introduced and discussed in Hypotheses 1 and 2. The first two properties define the common core of the fictional processing: they represent the crucial and indispensable conditions of the fictional experience. As such, they are subject to a small amount of variability. The second two properties contribute to determine the specificity of the fictional experience as distinct from nonfictional and nonnarrative experiences. However, they are subject to a large amount of variability and the fictional experience can also occur without them, at least in principle.
This large variability of transportation and identification depends on two general principles of human mind: accessibility and the role of contextual and individual factors. How accessibility affects the fictional processing is introduced and discussed in Hypotheses 3 and 4. These hypotheses focus on two key variables: epistemic vigilance and defamiliarization. On the one hand, epistemic vigilance contributes to determine different kinds of fictional experiences. On the other hand, defamiliarization contributes to determine the specificity of the more demanding fictional experience. Hypotheses 5 and 6 introduce and discuss the interaction of motivation and opportunity, and the role of the main individual variables, namely education, expertise, familiarity with subject matter, and personality traits, in particular curiosity and empathy. The review highlights the dynamical relationships among all these factors, showing that very different experiences of fiction emerge from their interactions.
The review discusses the prominent models of narrative persuasion, that is the Transportation Imagery Model (Green and Brock 2000), the Narrative Engagement Model (Busselle and Bilandzic 2008), and the Extended Elaboration Likelihood Model (Slater and Rouner 2002). In particular, the review is inspired to the dual-system theories (Stanovich and West 2000), and more specifically to some aspects of the elaboration likelihood model (ELM—Petty and Wegener 1999) and its extended version (Slater and Rouner 2002). It is well known that the predictions of the ELM change in relevant ways in the domain of not explicit persuasive messages. However, the idea of an elaboration continuum, the dual-route perspective to learning, and the multi-process nature of persuasion represent key concepts to understand the impact of fiction. In particular, in analogy with the ELM, the main construct of the present review is that the fictional processing may have very different realizations along an elaboration continuum. At the low end of the continuum, contents are highly accessible; information scrutiny is reduced; epistemic vigilance is low; and the experience of fiction is easy and fluent. At the high end of the continuum, contents are not immediately accessible as a consequence of deviations, distortions, ambiguities, uncertainties, and gaps; information scrutiny is strengthened; epistemic vigilance is high; and the experience of fiction is dynamically fluent: it starts from a state of uncertainty and reaches a subsequent state of increased predictability.
According to the ELM, at the low end of the continuum peripheral processing is constituted by very simple reactive mechanisms (such as mere exposure, priming, conditioning, and familiarity) and automated heuristics (such as the reliability of the author/narrator and the identification with characters). At the high end of the continuum extended, systematic, and deliberate cognitive control is engaged. Each experience of fiction represents a specific point along the continuum, determined by the fictional representations’ properties and quality as well as by the subjects’ motivation, personality, and expertise.
The common core of fiction: simulation and believability
Whatever their positions along the elaboration continuum, all the experiences of fiction share a specific processing style based on two common intertwined properties: simulation and believability.
Simulation is a basic human capacity and constitutes a unifying computation principle across diverse processes in the brain. It provides the reenactment of perceptual, motor, and introspective states acquired during the previous experience with the world (Barsalou 2008). Imagination can take very different forms and it does not permit a simple taxonomy (Gendler 2011). However, according to a widespread consensus, one of the main uses of the term considers imagination as a specific form of simulation (Currie 1995). Precisely, imagination provides not referentially constrained simulations. It can reenacts sensations, perceptions, beliefs, desires, emotions, objects, situations, and even global experiences, yet as imaginings they are invitations to imagine without real-world reference and factual truth conditions. To imagine something is not to be committed to its truth: imaginings are referentially void and do not have referential force (Schaeffer 1999). In the domain of fiction individuals know very well that simulations are decoupled from the actual state of affairs and that real-world truth conditions are irrelevant. They accept a kind of “fictional agreement”, based on what Coleridge defined as a “willing suspension of disbelief”. In this perspective works of fiction are guides to make-believe (Walton 1990; Currie 1990). They provide a safe arena in which individuals can vicariously experience emotions without real-world consequences and the need of self-protection (Menninghaus et al. 2017).
Fictional simulations represent a subset of narratives: they tend to be written in the narrative mode, that can be linguistic or iconic (Bruner 1986; Fludernick 2006). A narrative is a meaningful and temporally structured sequence of causally related events. It requires the presence of at least one change in a state of affairs: actions or events are in chronological order and stand in some kind of causal relation to one another (Herman 2002; Fludernick 2006). Narrative typically has an identifiable structure with a clear beginning, middle, and end during which some issue is brought to a state of completion and contains implicit or explicit messages about the topic addressed (Hynyard and Kreuter 2007). The “story” refers to a set of completed events which is reported through a narrative (Green and Brock 2000).
According to the widely accepted model of Oatley (2016), Oatley and Djikic (2017) and Oatley et al. (2018), the typical subject matter of the fictional narratives is selves and their interactions in the social world. They take into account complexes made up of multiple interacting processes. Emergent properties, that cannot be predicted in advance from the low-level interaction, arise from these processes. Fictional simulations describe the emergence of such possibilities, thus providing concrete simulative experiences (Mar and Oatley 2008; Mar 2018).
However, when imagination produces fictional simulations, it is not unconstrained: although fictional worlds are autonomous and self-contained, they must be plausible and convincing (Currie 2016). To be understood and appreciated, fictional simulations require believability (verisimilitude, plausibility, truthlikeness). In line with the Narrative Engagement Model and the central concept of perceived realism (Zwan 2004; Busselle and Bilandzic 2008), believability has two directions and can have different degrees. On the one hand, recipients judge the (degree of the) internal coherence and logic of fictional narrative, alongside with the truth of the events within the fictional context (internal, narrative consistency). On the other hand, subjects judge the (degree of the) truth of fictional events compared to real-life events by asking themselves whether certain events of the fictional narrative would be likely to happen in reality. However, it is important to stress that in my view believability is not the same of realism and factuality. Even completely unrealistic fictions may be plausible to the extent that recipients use ordinary beliefs to understand scenarios, plots, and characters. For instance, in genres like fantasy and science fiction characters generally adopt beliefs, rules, and standards derived from ordinary folk-psychology to cope with the unrealistic scenarios. On this basis recipients can judge the degree to which the causal sequences of the fictional plot is internally consistent and externally plausible. From this latter point of view, they judge whether or not the narrative could possibly happen in a future/potential/imagined (at present non-factual) real-life situation.
This point is clearly explained by the pivotal mechanism of believability. According to Currie and Ravenscroft (2002) this mechanism is substitutability. Fictional representations have the same code, format, and contents of ordinary ones and therefore they can be substitute for, and hence capable of simulating, ordinary representations. They can be inserted into ordinary inferential and reasoning mechanisms which will run off-line and process parallel representations much the same way. In this perspective fictional representations and ordinary ones are isomorphic to the extent that they share similar (but not identical: Nichols 2006; Galgut 2014) causal powers on subsequent processing. Therefore, the understanding of fiction is governed by the same rational rules that guide the understanding of the actual world. Fiction duplicates reality: if it is reasonable to believe that p in the ordinary context, it is also reasonable to imagine that p in the fictional context. As a consequence of substitutability and isomorphism, individuals can engage with fiction by using the same ordinary processes activated by real contexts.
Neuroscientific evidence significantly supports the hypothesis that fiction represents a simulation that promotes the exploration of opportunities and possibilities especially in the social domain. The crucial study of Altmann et al. (2012) investigates the fact/fiction distinction during the reading of simple stories. During the fMRI experiment, participants read a narrative that was presented for 20 s. The context label, that defines the narrative as either factual (real) or fictional (invented), was changed for different groups of participants. The different contextual labels were associated with the activation of different neural patterns. Both labels activated processes of imagination. However, they reflected different levels of simulation. The neural pattern associated with the factual context involved a broader sense of simulation, perceived as the inner representation of actions. Comprehension strategies were guided towards the understanding of what happened in the story: participants gathered information to update their world knowledge. Reading fiction involved a narrower sense of simulation, conceived as an imaginative construction of hypothetical scenarios. The engaged neural network corresponded to a constructive simulation of what might have happened or could happen. Thus, reading fictional texts exceeds the mere information gathering. “Readers perceive the events in fictional stories as the likelihood that something might have been, which leads to an active simulation of events—similar to the simulation of a possible past or a possible future” (Altmann et al. 2012, p. 26).
In line with the study of Altman et al. (2012) and Tamir et al. (2016) found that reading fiction and social cognition both recruited the default network, which supports our capacity to simulate experience outside of the ‘here-and-now’, such as thinking about the future and the past, mentally constructing places and spaces, imagining hypothetical events, and simulating others’ mental states. The quantitative meta-analysis of Mar (2011) supports the hypothesis that the understanding of fiction engages isomorphic processes to those activated in real-world circumstances. The study found that story processing overlapped with many key brain areas that support our ability to mentalize, the so-called mentalizing network.
The prototype of fiction: transportation and identification
The common core of fiction can have different realizations as a consequence of the interactions among several variables. Transportation and identification represent the key variables that contribute to determine the specificity of the fictional simulations as distinct from nonfictional and nonnarrative experiences.
According to the previous hypothesis, fictional simulations tend to be written in the narrative mode. Like every narrative, they can activate transportation and identification. From this point of view, fictional narratives typically do not present overtly persuasive messages associated with issue-related arguments and truth claims. On the contrary, according to the Transportation Imagery Model they typically activate transportation, a vicarious experience of the imagined events based on “a convergent process, where all the mental systems and capacities become focused on events occurring in the narrative” (Green and Brock 2000, p. 701). Transportation represents a mental journey into the imagined world of the fictional narrative and occurs when an individual departs from the real world, is fully engaged in the story, experiences high imagery, and is emotionally struck by the story (Gerrig 1993; Green and Brock 2002).
A large body of evidence suggests that transportation occurs across media (novels, films, television shows, and video games) and represents the key mechanism of narrative persuasion—the impact of narratives on beliefs, attitudes, and behaviors. In the typical experimental design, participants read a short story or a lengthy excerpt from a popular novel. After reading they rate their transportation into the story on the basis of specific scales created to assess the degree of individuals’ immersion into a narrative, finally they respond to a series of belief statements related to assertions in the story. The results always indicate that transportation increases endorsement of story-consistent beliefs: more transported individuals show more story-consistent beliefs (Green and Brock 2000a, b; Green 2004). This effect occurs regardless of whether the narrative is labeled as a fact or a fiction (Green and Clark 2012).
There is a general consensus that transportation plays a pivotal role in narrative persuasion by reducing disbeliefs, cognitive resistance, and counterarguing—the generation of thoughts that dispute the persuasive content (Green and Brock 2002; Dal Cin et al. 2004; Busselle and Bilandzic 2008; Green and Clark 2012). Even when a persuasive message is explicitly presented, transported individuals may be unwilling to actively dispute or reject it, because they do not want to disrupt their enjoyment by breaking out the narrative world to critique points made in the story. Moreover, transportation typically facilitates connections with characters, leads to vivid mental imagery and rich mental pictures, increases the perceived realism of the narrative, makes the narrative events more like personal experience, and triggers strong emotional responses, other key factors that boost the believability of the messages and therefore facilitate persuasion (Green 2004).
According to the Extended Elaboration Likelihood Model, also identification with characters is a key mediator of the narrative persuasion and, like transportation, reduces counterarguing (Slater and Rouner 2002; Dal Cin et al. 2004). Identification typically represents an imaginative process through which a recipient temporarily suspends the self and takes on the perspective of a fictional character, participating in the events of the story. This complex process engages different dimensions, any one of which may or may not occur in any given instance: the empathic dimension (sharing emotions with character), the cognitive dimension (sharing character’s perspective), the motivational dimension (sharing character’s goals and attitudes), and absorption (involving the loss of self-awareness during exposure). This latter dimension is different from transportation because it is directed at one particular character rather than at a storyline overall (Cohen 2006).
Identification is only one kind of involvement with characters, that represents a broad category made up of different phenomena (Moyer-Gusé 2008). A related construct is “wishful identification”, that occurs when a viewer wants to be like and emulate the character (Giles 2002). In the other forms of involvement with characters the recipient adopts a different perspective: s/he is entirely her/himself and make judgment about the character. Homophily refers to the perceived degree to which an individual perceives that s/he is similar to a character. Liking simply refers to a positive evaluation of a character. Parasocial interaction is a one-side pseudorelationship with the character, seen as part of her/his social world, like a friend. Anyway, whatever the specific realization of identification and involvement, participation in the story and the adoption of character’s beliefs, emotions, goals, and attitudes generally lead recipients to an affinity with characters, by prompting an attitude towards them based on liking and empathy (Kotovych et al. 2011). Furthermore, identification stimulates greater cognitive elaboration and a more complex reflexive process of thinking about the topics dealt with in the narrative (Iguarta 2010).
If there is a large body of evidence showing the effects of fictional narrative on real-world beliefs through transportation, the impact of fiction—precisely of a one-time exposure—on social cognition, in particular on mind reading, cognitive and affective empathy, and prosocial attitude and behavior, is more controversial. Mind reading is the everyday ability to make sense of the behaviors of others: it allows us to understand others as intentional agents and to engage in shared activities (Tomasello 2015). There is a widespread agreement in distinguishing two types of mind reading (de Vignemont 2009). The first one is very simple; it is based on automatic, unconscious, prereflexive mechanisms; it does not involve inferences and propositional contents. The basis for such mind reading is automatic “mirroring” and unmediated resonance for other’s emotions, feelings, intentions, and actions. In contrast, high level mind reading engages the complex use of imagination as a constructive process that allows individual to assume the other’s perspective. This is an imagination-driven simulation based on “perspective-taking” and aimed at removing imaginatively any relevant disparities between the simulator and the target. This kind of mind reading based on perspective-taking can be exclusively cognitive: an individual represents and understands the mental states of another person, without being emotionally involved (“cognitive empathy”). When an individual shares an isomorphic affective and emotional state with another person and she/he is conscious that this other person is the source of the affective state, this mind reading becomes “affective empathy” (de Vignemont and Singer 2006).
The pivotal role of transportation in moderating the effects of fiction on social skills is suggested by the study of Bal and Veltkamp (2013) and by the study of Johnson (2012). The study of Bal and Veltkamp (2013) investigates affective empathy, that is emotional sharing. In the fiction condition, participants read the first part of a Sherlock Holmes short story. In study 2 they read the first chapter from Blindness by Saramago in order to ascertain whether the effects of fiction hold across genres. In the control condition, participants read a selection of emotionally involving articles from newspapers. Participants worked remotely and read the story on line via computer. Emotional transportation was measured directly after reading the text, using the scale of Busselle and Bilandzic (2009). Empathy was assessed directly before the experiment, directly after reading the text, and one week after the experiment using the emphatic concern subscale of the Interpersonal Reactivity Index (IRI). Self-reported emotional empathy increased over a period of one week for those who had read a fictional story (this is called an “absolute sleeper effect”), but only when they were emotionally transported into the story (even if in study 2, the increase of empathy was not significant). The reverse occurs when the recipient does not become transported: then the recipients becomes less empathic.
A similar moderating role of transportation has been found also for the effects of fiction on the behavioral dispositions towards real-world targets. In the study of Johnson (2012), participants read a story written specifically for the study designed to arouse compassionate feelings for the characters. Participants rated the extent to which they had experienced six emotions while reading the story, including compassion, sympathy, soft-heartedness, tenderness, sentiment, and warmth. A helping behavior was observed where the experimenter “accidentally” drops pens within sight of the participants and then records whether the participants help pick up the pens. Results showed that more transported participants reported significantly higher affective empathy for the characters and were significantly more likely to help the experimenter pick up the pens. Certainly, this effect refers to a trivial and short-lived helping behavior. Moreover, the study of Koopman (2015a) on medium-term prosocial effects found different results. The participants were exposed to three kinds of texts: literary (with a high degree of foregrounding of language and style), “life” narrative, and non-narrative instructional texts. For the reading of each of the two narrative texts, the participants were divided into two groups: one was told that the text was fiction and the other that it was non-fiction. The effect on prosocial behavior was measured by the willingness to donate a small amount of money. Results were inconclusive, because only few donated. Moreover, life narrative was more effective than literary narrative and it made no difference whether or not the participants were told that they were reading fiction.
The evidence concerning the effect of fiction and transportation on cognitive empathy is even more controversial. Kidd and Castano (2013) reported that a brief, one-time exposure to literary fiction can immediately enhance cognitive empathy. Cognitive empathy was measured using the “Reading the Mind in the Eyes Test” (RMET), a performance-based measure. The RMET shows respondents pictures of actor’s eye-region and asks which of the four possible mental states is associated to that people (Baron-Cohen et al. 2001). Respondents choose among four terms to indicate what each photographed person was thinking and feeling. The correct responses indicate the ability to understand and pair mental-state terms with static non-verbal cues. In the study of Kidd and Castano (2013) participants were randomly assigned to read short excerpts of literary or popular fiction texts, expository non-fictional texts and no texts. Participants in the literary fiction condition showed higher scores on RMET compared to those in the other conditions.
Various replications and extensions of Kidd and Castano (2013) found similar results. Adopting a within-subjects design, Black and Barnes (2015a) replicated the results showing the moderating role of transportation. Black and Barnes (2015b) and Bormann and Greitemeyer (2015) found similar results on RMET using other media, respectively award-winning TV dramas vs. documentary TV programs, and video games, introduced in a narrative way vs. in a technical way. Pino and Mazza (2016) replicate the findings using a large battery of mentalizing and empathy tests during the pre and post-reading phase. However, the replications of Panero et al. (2016) and Samur et al. (2017) did not find the results of Kidd and Castano (2013). These two studies had a higher statistical power because they had a larger sample than Kidd and Castano (see also the controversial results in Kidd and Castano 2018).
However, these mixed and hesitant findings do not discredit the persuasive impact of fiction on real world beliefs and attitudes. On the contrary, they indicate the fragility of the priming paradigm, focused on the effect of one brief exposure to a short excerpt of the priming stimulus (Panero et al. 2016; Samur et al. 2017). This is a not ecological paradigm, very far from the real world conditions of processing, understanding, and interpreting fiction (Koopman and Hakemulder 2015; Samur et al. 2017; Barnes 2017).
Persuasion by fiction through heuristics
When the experience of fiction is based on high accessible contents, transportation tends to be associated with low epistemic vigilance; as a consequence of low epistemic vigilance, persuasion is based on a stimulus-driven, involuntary, and implicit transmission of information and it may be far from learning.
According to Gerrig and Wenzel (2015) all the narrative experiences involve active and complex inferential processes that provide the groundwork for the participatory responses described in the previous hypothesis. These inferences concern the information that is not explicitly stated in the text. They integrate the text with prior knowledge; connect present and past; anticipate what the recipient expects to follow; cope with the large number of gaps between what the author/narrator knows and what the recipient knows; attribute nonliteral meaning to the text. Moreover, even literary novices are capable of engaging in elaborative processing for inferring interpretations (Burkett and Goldman 2016). However, epistemic vigilance has a deep impact on these inferential processes and contributes to determine very different experiences of fiction. According to Sperber et al. (2010), epistemic vigilance is a critical stance towards the communicated information. It involves several cognitive mechanisms that work in parallel to ensure that communication remains advantageous despite the risk of being accidentally or intentionally misinformed. These mechanisms contrast distraction, identify the relevance of what is communicated, assess the reliability of the source and the content, yield a variety of epistemic attitudes (acceptance, doubt, rejection). Even at a very early age children do not treat all communicated information as equally reliable. For instance, there is evidence that at 16 months they notice when a familiar word is inappropriately used. However, the transition to epistemic vigilance seems to occur around the age at which children succeed in passing standard false belief task. This indicates that the improved capacity for epistemic vigilance involves a major development in executive function abilities.
The experience of fiction can be based on highly accessible contents. Especially popular fiction often presents familiar prototypes, that is stereotypical, fixed, and oversimplified scenarios, characters, lines of stories (Barnes 2017). These contents easily come to mind and prompt fast and automatic operations. Associated with transportation, this routinized processing tends to lower epistemic vigilance. As we have seen in the previous hypothesis, transported recipients are fully identified with the fictional world and, under this special experiential state, an active construction of disbelief that can lead to the rejection of belief-incongruent information is unlike. In this condition, transportation reduces and blocks cognitive resistance, negative responding and counterarguing (Dal Cin et al. 2004). Persuasion occurs without conscious control and careful scrutiny: recipients tend to automatically accept the communicated information, without attending to the reasons for accepting it. They are largely trustful and minimally vigilant.
Elementary reactions and automated heuristics, typical of the peripheral processing (Petty and Wegener 1999), are determinants of persuasions. Elementary reactions such as priming, classical conditioning, and familiarity do not rely on controlled and issue-relevant thinking. The most relevant heuristics are the identification with characters and the assumption of the reliability of the source. As we have seen in the previous hypothesis, identification reduces disbeliefs and counterarguing. In the same vein, recipients spontaneously have a high degree of trust in the author/narrator as a reliable informant, automatically assumed as competent and benevolent. They use the credibility of the author/narrator as an easily available discounting cue which leads them to reject any belief-incongruent information (Appel and Richter 2007; Appel and Mara 2013).
Epistemic vigilance copes with “the risk of being accidentally or intentionally misinformed” (Sperber et al. 2010, p. 360). Accordingly, when epistemic vigilance is low and there is not cognitive control, persuasion is largely unconscious, implicit, and stimulus-driven. As a consequence of low epistemic vigilance, the impact of fiction is not necessarily a positive benefit, that is an improvement from a worse belief/attitude to a better one. Individuals can easily acquire false beliefs and negative attitudes.
A large body of evidence supports all the main points of Hypothesis 3. Many findings confirm the crucial role of the peripheral factors (Marsh et al. 2003; Appel and Richter 2007; Appel and Mara 2013; Richter et al. 2014). For instance, the study of Richter et al. (2014) shows the connection between transportation and priming. The study focused on self-reported femininity and found that a fictional story can alter the self-concept. The experimental group read a story featuring a protagonist with a traditional gender role. The effect was found only among readers who were more deeply transported. Interestingly, transportation was positively associated with self-reported femininity in both the experimental and the control condition, based on a story without information related to a feminine gender role. According to the authors, the simple fact of transportation typically involves perspective-taking and empathy, experiences arguably related to the femininity concept. So, in line with classical research on priming (Bargh 2006), the simple fact of transportation unconsciously primed subsequent changes in the self-concept without any conscious control and intentionality.
The crucial role of the law epistemic vigilance, associated with a high condition of transportation, is supported by the study of Appel and Mara (2013). This study investigates the impact of a fictional story on fuel-efficient driving. In this study the trustworthiness of the character that delivered the information on fuel-efficient driving was manipulated. Behavioral intentions were assessed briefly after exposure. Transportation served as a moderator variable. Average green driving intentions were higher in the trustworthy group than in the untrustworthy group. Low character trustworthiness was particularly detrimental to story-consistent intentions for recipients who were not deeply transported into the story world. The influence of the character low in trustworthiness was higher among more transported recipients and it increased with the recipients’ transportation. Strong transported participants were equally persuaded by both trustworthiness versions. This finding shows “a lack of any critical evaluation of the story’s main message under high narrative presence” (Appel and Mara 2013, p. 23): transported individuals are particularly unlikely to engage in the process of scrutinizing and rejecting information.
The crucial role of identification, conceived as an automated heuristic, is supported by Vezzali et al. (2015). Study 1 assesses whether Harry Potter, the popular fantasy novel, can be used to improve the attitude towards stigmatized groups on the basis of the identification with the main positive character. The character of Harry entails positive personal and social values, while Voldemort, the main negative character, believes that power should be held only by “pure blood” wizards. Experiment 1 was conducted among Italian elementary school children. In the first phase, children were administrated a questionnaire assessing their attitudes towards immigrants. Passages from Harry Potter were read and then discussed with a research assistant once a week for six consecutive weeks. Selected passages were related (experimental condition) or unrelated (control condition) to the theme of prejudice. After reading, children took part in a group discussion led by the researcher focusing on what they had just read. In the experimental condition, the discussion focused on the prejudicial acts committed against stigmatized groups, on their consequences, and on the role of Harry Potter. One week after the last session, children were administrated a questionnaire with the dependent variables, the identification with Harry Potter and Voldemort. The results showed improved attitudes towards immigrants among those children who identified with Harry Potter, while the identification with Voldemort was associated with more negative out-group attitudes. Thus, this study indicates that fiction, through identification, influences real-world attitudes. The net effect of reading fiction is still unclear: it is impossible to differentiate this effect from that resulting from the discussion guided by the researcher. However, considering the age of participants, it is clear that the process of identification activated in this study is far from the complex self-reflection that will be discussed in the next hypothesis.
Finally, the study of Marsh et al. (2003) and the study of Appel and Ritcher (2007) show that fiction can change real-world beliefs on the basis of familiarity. In the study of Marsh et al. (2003) participants read short fictional stories that contain peripheral references to information about the real world and took a general knowledge test. Some of the fact in the story correspond to questions on this test. More questions were properly answered after having read the correct facts in the stories. In study 2 the stories included correct and incorrect factual information. The results show a strong misinformation effect with the production of misinformation increasing dramatically only when participants read misinformation in the stories.
The study of Appel and Ritcher (2007) uses the method of agreement extremity: to measure the effects on beliefs, researchers assess the participants’ agreement with the propositions at stake. The persuasive effects of a fictional story were measured either immediately or two weeks after reading with this method. The results showed that there was a significant short-term persuasive influence of false information: false assertions about particular subject matters reduced the extent at which participants endorsed true beliefs with respect to those subject matters. This influence was higher upon a delay of two weeks compared to immediately after the reading, resulting in what is called an absolute sleeper effect. Thus, taken together the studies of Marsh et al. (2003) and Appell and Ritcher (2007) suggest that, when epistemic vigilance is low, individuals can derive false real-world beliefs from fiction without any conscious control. Moreover, they suggest that this effect results in a lasting belief change—in contrast to the prediction of ELM, according to which only attention to message’s arguments has the potential for lasting change.
Persuasion by fiction through reflection
When the experience of fiction is based on defamiliarization, transportation tends to be associated with high epistemic vigilance and reflection; as a consequence of high epistemic vigilance and reflection, persuasion requires a demanding process of effortful scrutiny and involves a flexible reasoning on the substantive merits of the contents.
Defamiliarization contributes to determine the most demanding experience of fiction. The best example of a fictional experience of defamiliarization is supported by literature. Literature is a subset of narrative that partly overlaps fiction. Fiction includes many non-literary texts and literature includes many non-fictional texts. Moreover, fiction is a descriptive concept, while literature depends on aesthetic evaluation concerning originality and stylistic quality.
Even if the social mechanisms that make these value judgments appear change over time and are not always clearly linkable to text features, the scientific approach to literature tries to operationalize the ‘literariness’ by referring it to the concept of “foregrounding” (Mukařovský 1976). This concept indicates the unusualness and unconventionality of language, contents, and ways of telling a story. Foregrounding can be phonetic, grammatical, and semantic. Obviously, it is not exclusive to literature: foregrounding features also appear in nursery rhyme, slogans, proverbs, and so on. As an experience of deviation, foregrounding causes de-automatization and “defamiliarization” (Shklovsky 1965): it could lead readers to become unsettled and to start looking at thing differently (Miall and Kuiken 1994; Van Peer et al. 2007; Hakemulder and van Peer 2015).
Foregrounding activates an original implementation of the fictional processing with a specific realization of simulation and believability, inference and epistemic vigilance, transportation and identification. Simulation and believability cannot be based only on recreative imagination, that is on the reenactment of well-rehearsed simulative patterns previously entrenched in memory (Currie 2001). Creative imagination comes into play and works creatively on existing simulative patterns to match deviations and distortions from what is normally expected (Boden 2004). Recipients raise their vigilance and actively check the plausibility and acceptability of their interpretative inferences (Origgi 2013). Associated with active vigilance, the inferential process becomes “reasoning”—“a form of inference which involves attending to the reasons for accepting some conclusions. Reasoning, so understood, involves reflection, and contrasts with intuitive forms of inference where we arrive at a conclusion without attending to reasons for accepting it” (Sperber et al. 2010, p. 377). Recipients engage in “deep reading” (Hall et al. 2015), i.e. slowed and thoughtful reading. They re-oriented their attention, moving from the source to the contents: the simulation is approached as an insightful model independent from the author, endowed with an intrinsic credibility (Currie 2016). Under the condition of high epistemic vigilance, transportation is far from being an easy experience of passive and quick consumption and identification is other than an automated heuristic. They co-occur with active sense-making, critical assessment, and rational reflection. Transportation becomes an “imaginative engagement” (Barnes 2017): recipients actively and imaginatively cocreate with the author the narrative, “by filling in gaps, puzzling over interpretations, fleshing out what is written, or other imputing meaning onto the page that extends beyond what is written” (Barnes 2017, p. 3). The heuristic of identification becomes a complex and extended process of self-reference that may lead to a renewed self-understanding.
At present the main source of evidence concerning foregrounding comes from the phenomenological—albeit rigorous and numerically aided—studies based on multiple content analysis of readers’ responses (Miall and Kuiken 1999, 2002; Kuiken et al. 2004; Sikora et al. 2010;). First, several studies showed that deviations and distortions in grammatical sentences, in the use of metaphors, and in the order of the story reduce the processing fluency and require longer reading times (Miall and Kuiken 1994; Hakemulder 2007). However, the data-driven qualitative study of Bálint et al. (2016), based on the analysis of in-depth semi-structured interviews, suggests not only that foregrounding and transportation may co-occur in the same experience, but also that foregrounding may deepen transportation.
Second, many studies indicate that the feelings specific to literary reading, called “self-modifying feelings”, play a critical role as the vehicle of a deeper or changed self-understanding, typically based on the connection that the readers perceive between their personal experiences and the images used in the text. The so-called “enactive reader” progressively turns into an affective theme across striking passages according to this sequence. Foregrounding and the state of uncertainty evoke a feeling arising from the text with some personal relevance. As the vehicle of interpretation, the feeling initiates a process in which the existing schemata become recontextualized: the personal experience offers a new interpretative context. In this way, the feeling is used to find a convergence between the character’s situation and that of the reader to achieve a generalization that includes the reader. Within this generalization, the reader compares the characters with her/himself/herself to capture any similarity. The reader identifies with the character as a member of the same inclusive and ad hoc created class. The recontextualization of the schemata leads to new insights. The affinity between the reader and the text simultaneously prompts openness to a different understanding not only of the text, but also of the reader. In this perspective, the foregrounded stylistic and narrative aspects may not only have a direct defamiliarization effect, but also initiate a form of reflexivity, in which readers may appropriate the text’s figurative forms, by borrowing its metaphors, in reflective reference to the aspects of their own lives. This is a very complex process of identification, very different from the automated heuristic of identification described in Hypothesis 3.
Finally, the qualitative-quantitative study of Koopman (2015b) showed that, against expectations based on previous studies (Miall and Kuiken 2002; Sikora et al. 2010), for the immediate reflection condition, the expository texts evoked more personal thoughts, while, as expected, for the deferred reflection conditions respondents had thought back to the narrative more frequently than to the expository. This supports the idea that the understanding of literature exceeds the act of mere reading and is temporally extended. Individuals think back and mentally relive the story that they have read and link the information from fictional narratives with their backgrounds and daily encounters.
Motivation and opportunity
The processing and the effects of fiction are mediated by the interaction of opportunity and motivation.
In line with the dual-system theories (Fazio and Towles-Schwen 1999), it is generally assumed that the default processing mode keeps cognitive effort to the required minimum, especially when motivation is low. This means that some motivating goal is necessary to induce individuals to be accurate and engage in effortful reasoning. This motivation generally stems from the perceived costliness of a judgmental mistake: fear of invalidity motivates careful scrutiny. The personal costliness of accuracy largely derives from the “ideal self image”, usually conceived as a hierarchical structure with a collection of ideal traits and concepts that the individual aspire to Pelowski et al. (2017). Moving from top to bottom, this branches into general schema for pursuing ideal traits and concepts, which are themselves divided into more and more local expectations for specific behavior. For example, whereas for many individuals art may not play a core role, art professionals may see art as highly personal relevant (“be someone connected to art” > “do understand artworks” > “find meaning in a particular artwork”) (Pelowski 2015).
When individuals are transported they deliberately accept the fictional agreement and suspend disbeliefs and counterarguing. To the extent that contents are highly accessible and believability and isomorphism are fluent and easy, the experience of fiction tends to be a quick and passive consumption. Individuals do not have the general motivation to strengthen epistemic vigilance, to engage in effortful scrutiny, and to reject belief-incongruent information. At this low level of information scrutiny, elementary reactions and automated heuristics represent key determinants of persuasion.
Obviously, the motivation to engage in effortful processing is not sufficient per se. Fictional representations must provide opportunities to undertake the sort of imaginatively engagement and reflection that occur in the case of literature: striking distortions, challenging gaps, implicit meanings, unfamiliar metaphors, multidimensional and hard to predict “round” characters, and so on. Under these high-elaboration conditions, deliberate and analytic processes are engaged and heuristics and arguments are subjected to careful scrutiny.
The studies of Dijkic and collogues support the pivotal role of the quality of the fictional prompt. In the study of Djikic et al. (2009), after completing a series of questionnaires, including the most widely used model to describe how a person relates with the world (the Big-Five Inventory—BFI) and an emotion checklist, the experimental group read the The Lady with the Dog short story by Chekhov, while the control group read a comparison text that had the same content as the story, but was documentary in form. The length, readability, complexity, and interest level of the comparison text were controlled. Participants then completed again the BFI and the emotion checklist. Participants who had read the Chekhov story scored significantly higher trait change and emotion change than those who had read the comparison text. These changes, each in her or his own way, were small, but statistically significant. Djikic et al. (2012) replicated this study. Participants were randomly assigned to read either one of eight literary short stories or one of eight literary essays. Compared with those who read a work (short story or essay) they judged not to be artistic, those who read a text that they judged to be artistic changed their personalities by small amounts. Therefore, in this study literature was responsible for change, and not fiction vs. nonfiction.
However, it is important to stress that at present the correlation studies do not determine whether or not the impact of fiction on social cognition is specific to literary fiction. These studies are based on the “Author Recognition Test” (ART), The ART asks respondents to tick from a list of names those relating to authors (Acheson et al. 2008). It acts as an adequate measure of lifetime exposure to literature. A modified version includes a strict categorization of the names into two mutually exclusive categories of fiction and non-fiction. With this modification, the people’s reading of fictional and non-fictional texts can be estimated. The association between overall ART and RMET performance has been replicated in each study many times. However, there is no consensus on whether this result is specific to literary fiction (Kidd and Castano 2016) or if it is generalized to genre/popular fiction, in particular to romance (Fong et al. 2013).
Moreover it is also possible that, as a consequence of their particular self-image, individuals engage in deep reading when they approach popular fiction. From this point of view, empirical research on self-identified fans shows that they are personally motivated to engage with popular fiction with a type of processing that goes far beyond the passive consumption of the text. Their processing parallels the type of imaginatively engagement required by literary fiction: they engage in in-depth discussion with a high level of scrutiny and critical analysis; they approach the text metaphorically, not just at the sentence level; they use subtle cues to make complicated inferences about characters’ mental states; when they write fanfiction, they aims at filling narrative gaps and making emotional subtext textual; as a consequence of the identification with the favorite characters and the adoption of their perspective outside the text itself, readers personalize the text and draw personal meaning from it after reading (Barnes 2017).
The processing and the effects of fiction are mediated by a set of individual variables capable of influencing the crucial factors of transportation and identification; the most relevant variables are education, expertise, experience with subject matter, and trait personality, in particular curiosity and empathy.
When the fictional narrative overtly communicates factual persuasive messages, recipients with a high level of education can reduce and even block transportation and identification because these are perceived as not appropriate to the communicated information. For instance, this effect occurs in the study of Kreuter et al. (2007) that investigates the impact of fictional communication on perceived social norms, in particular concerning cervical cancer and the attitude towards Pap test.
Many studies investigate the role of literary and art expertise, suggesting that experts and non-experts differ in how they literally experience, understand, and evaluate art (Locher et al. 2010; Silvia 2013). Experts are less responsive to the direct affective valence of the fictional representation than are non-experts; they emphasize the structural, stylistic, and formal properties more than content, craftsmanship, and level of realism (Leder et al. 2014). Literary novices can move beyond textbase and situation level representations and engage in literary interpretation, providing interpretations for symbols and abstracting a larger message from the text. However, they do not engage in the same literary reasoning as experts. In particular, they do not generalize: they do not interpret the story as a whole and do not connect it to their conceptions of the human condition (Burkett and Goldman 2016).
Differences in familiarity can prompt different processes of elaboration: when recipients can draw on their prior experience, they are facilitated to become absorbed and identified (Green 2004). From this point of view, the study of Koopman (2016) shows that the role of familiarity is more relevant than foregrounding in causing affective and emotional responses.
Finally, two personality factors, curiosity and dispositional empathy, have a crucial role in determining the effect of fiction. Regardless of expertise, curiosity can facilitate a greater engagement and a deeper processing of the fictional narratives, especially if these narratives are novel and original (Fayn et al. 2015). Dispositional empathy can facilitate identification and emotional responses correlated with characters’ emotions. The study of Gernot et al. (2017) found that participants with higher disposition for emotion contagion, namely the tendency to automatically pick up, mirror, and synchronize to emotions displayed by others, showed more congruent and more intense bodily reactions as measured by facial electromyography and skin conductance responses, and reported more intense aesthetic evaluations (higher being moved, valence, and interest), and also liked the artistic stimuli more (abstract and representational paintings). The recent study of Peretti (2018) shows the relationship between empathy and aesthetic abilities on the basis of the mediation effect of sleep.
Strength of effects induced by fiction
At low levels of the elaboration continuum, effects induced by fiction can strengthen their magnitude over time and represent a significant predictor of behavior as a consequence of high accessibility. At high levels of the elaboration continuum, fiction requires a temporally extended and systematic elaboration to cause idiosyncratic effects, specific to each individual.
The effects induced by fiction can be caused by both lifetime exposure and/or one-time exposure. The positive association between the amount of lifetime fiction exposure as measured by ART with RMET performance has been replicated in every study that used these tests. However, these results are only correlational: do people strengthen their mind reading capacity by reading fiction or are people with higher mind reading skills more attracted by fiction (Mar et al. 2006, 2009; Oatley 2016)? Moreover, the meta-analysis of Mumper and Gerrig (2017) found that this long-term effect was very small in magnitude.
According to the present review and in line with the ELM predictions, sometimes the effect of one-time fiction exposure occurs as a result of considerable reasoning about, and at other times it is associated with minimal reasoning. At the low end of the elaboration continuum elementary reactions and automated heuristics are relevant determinants of persuasion. Effects may be rather brief, limited to a short period, and weak. This is probably what happens in the study of Johnson (2012), in which participants seem to be unconsciously primed to prosocial behavior. Based on a similar priming paradigm, the study of Kidd and Castano (2013) found a small effect on cognitive empathy, corresponding to a 1-point difference on a 36-point scale.
However, in contrast to the ELM research and the prediction that only central and systematic processing, based on the attention to message’s arguments, has the potential for lasting belief and attitude change, there is crucial evidence suggesting that high accessibility facilitates the increase of the effect degree over time, resulting in medium-term effects. Appel and Richter (2007) found that the factual information derived from fiction can activate an absolute sleeper effect, and Bal and Veltkamp (2013) found that affective empathy induced by fiction can activate the same effect. Moreover, the strength of the effects induced by fiction may be very relevant in terms of prediction of behavior, as it is shown by the correlation studies on attitudes induced by exposure to movies smoking and alcohol abuse (Morgenstern et al. 2011; Dal Cin et al. 2012).
At the high end of the continuum, effects induced by fiction require a temporally extended process of elaboration in which individual’s personal background is activated and may be restructured on the basis of insights derived from fiction. The nature and magnitude of the effect are highly personal: they largely depend on how the face-to-face, sense-making interaction with the fictional representation engages recipient’s personal background and activates complex and demanding operations of reflection. As a consequence of personal engagement, the changes induced by fiction through reflection are not universal for all individuals. The qualitative-quantitative approach found that literary excerpts with a higher rate of phonetic, grammatical, and semantic foregrounding required longer reading times (Miall and Kuiken 1994); that reflection develops through a sequence of progressive transformations of a striking theme (Sikora et al. 2010); that literary texts enhance deferred reflection (Koopman 2015b); that only a small percentage of recipients (10%) reported changes and only sometimes (Miall and Kuiken 2002).
In the last decade there has been a great explosion of studies on fiction and its effects across scientific disciplines. However, there remain many controversial results and open questions (Koopman and Hakemulder 2015; Oatley 2016; Barnes 2017; Mar 2018). In this paper I address one of the most relevant limitations: the lack of an interdisciplinary integration. I put forth a theoretical-empirical review based on seven intertwined hypotheses. Each hypothesis connects different lines of research. Taken together, they integrate in a consistent theoretical framework the key mediating factors (transportation, identification, and perceived realism) of narrative persuasion identified by the prominent models and explain how these factors interact with two general properties of human mind, that is accessibility and personal and contextual variables.
According to these hypotheses, all the experiences of fiction share a specific processing style based on two common properties: simulation and believability. Transportation and identification are part of the prototype of the fictional experience and represent the key variables that contribute to determine the specificity of the fictional simulations as distinct from nonfictional and nonnarrative experiences. When the experience of fiction is based on high accessible contents, transportation tends to be associated with low epistemic vigilance; as a consequence, persuasion is based on stimulus-driven, involuntary, and implicit transmission of information and it may be far from having a positive impact. On the contrary, when the experience of fiction is based on defamiliarization, transportation tends to be associated with high epistemic vigilance and reflection; as a consequence, persuasion requires a demanding process of effortful scrutiny and involves flexible reasoning on the substantive merits of the contents. The processing and the effects of fiction are mediated by several contextual and individual factors: on the one hand, the interaction between objective opportunity and subjective motivation; on the other hand, the individual factors that can influence the experience of transportation and identification, i.e. education, expertise, familiarity with subject matter, and personality traits like curiosity and empathy. At low levels of the elaboration continuum, effects induced by fiction can strengthen their magnitude over time and represent a significant predictor of behavior as a consequence of high accessibility; at high levels of the elaboration continuum, fiction requires a temporally extended and systematic elaboration to cause idiosyncratic effects, specific to each individual.
- Appel, M., & Mara, M. (2013). The persuasive influence of a fictional character’s trustworthiness. Journal of Communication, 63, 912–932.Google Scholar
- Appel, M., & Richter, T. (2007). Persuasive effects of fictional narratives increase over time. Media Psychology, 10, 113–134.Google Scholar
- Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., & Plumb, I. (2001). The “Reading the Mind in the Eyes” test revised version: A study with normal adults, and adults with Asperger syndrome or high-functioning autism. Journal of Child Psychology and Psychiatry, 42, 241–251.PubMedCrossRefGoogle Scholar
- Bruner, J. (1986). Actual minds, possible worlds. Cambridge, MA: Harvard University Press.Google Scholar
- Cohen, J. (2006). Audience identification with media characters. In J. Bryant & P. Vorderer (Eds.), Psychology of reading, Mahwah (pp. 183–198). Mahwah, N.Y: Lawrence Erlbaum.Google Scholar
- Currie, G. (1995). Imagination as simulation: Aesthetics meets cognitive science. In M. Davies & T. Stone (Eds.), Folk psychology: The theory of mind debate (pp. 245–268). Oxford: Basil Blackwell.Google Scholar
- Currie, G. (2001). Imagination and make-believe. In B. Gaut & D. Lopes (Eds.), The Routledge companion to aesthetics (pp. 253–263). Abingdon: Routledge.Google Scholar
- Dal Cin, S., Zanna, M. P., & Fong, G. T. (2004). Narrative persuasion and overcoming resistance. In E. S. Knowles & J. A. Linn (Eds.), Resistance and persuasion (pp. 175–191). Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers.Google Scholar
- Fazio, R., & Towles-Schwen, T. (1999). The mode model of attitude-behavior processes. In S. Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology (pp. 97–116). New York: The Guilford Press.Google Scholar
- Fludernick, M. (2006). An introduction to narratology. New York, NY: Routledge.Google Scholar
- Galgut, E. (2014). Harnessing the imagination: The asymmetry of belief and make-believe. Contemporary Aesthetics, 12, 21.Google Scholar
- Gendler, T. (2011). Imagination, The Stanford encyclopedia of philosophy.Google Scholar
- Gernot, G., Pelowski, M., & Leder, H. (2017). Empathy, Einfühlung, and esthetic experience: The effect of emotion contagion on appreciation of representational and abstract art using fEMG and SCR. Cognitive Processing, 19, 147–165. https://doi.org/10.1007/s10339-017-0800-2. (Epub ahead of print).CrossRefPubMedGoogle Scholar
- Gerrig, R. J. (1993). Experiencing narrative worlds: On the psychological activities of reading. New York: Haven UP.Google Scholar
- Green, M. C., & Brock, T. C. (2002). In the mind’s eye: Transportation-imagery model of narrative persuasion. In M. C. Green, J. J. Strange, & T. C. Brock (Eds.), Narrative impact: Social and cognitive foundations (pp. 315–341). Mahwah, NJ: Erlbaum.Google Scholar
- Hakemulder, F., & van Peer, W. (2015). Empirical stylistics. In V. Sotirova (Ed.), A companion to stylistics (pp. 251–274). London: Continuum.Google Scholar
- Hall, M. P., O’Hare, A., Santavicca, N., & Jones, L. F. (2015). The power of deep reading and mindful literacy; An innovative approach in contemporary education. Innovación Educativa, 15, 49–60.Google Scholar
- Herman, D. (2002). Problems and possibilities of narrative. Nebraska: University of Nebraska Press.Google Scholar
- Iguarta, J. J. (2010). Identification with characters and narrative persuasion through fictional feature films. The European Journal of Communication Research, 35, 347–373.Google Scholar
- Kidd, D. C., & Castano, E. (2016). Different stories: How levels of familiarity with literary and genre fiction relate to mentalizing. Psychology of Aesthetics, Creativity, and the Arts., 11, 1–13.Google Scholar
- Kidd, D. C., & Castano, E. (2018). Reading literary fiction and theory of mind: Three preregistered replications and extension of Kidd and Castano (2013). Social Psychological and Personality Science, 20, 1–10.Google Scholar
- Koopman, E. M. E. (2015b). How texts about suffering trigger reflection: Genre, personal factors, and affective responses. Journal of Aesthetics, Creativity and the Arts, 9, 430–441.Google Scholar
- Mar, R. A., Oatley, K., Hirsh, J., dela Paz, J., & Peterson, J. B. (2006). Bookworms versus nerds: Exposure to fiction versus non-fiction, divergent associations, with social ability, and the simulation of fictional social worlds. Journal of Research in Personality, 40, 649–712.CrossRefGoogle Scholar
- Mukařovský, J. (1976). On poetic language. Lisse: De Ridder Press.Google Scholar
- Oatley, K., & Djikic, M. (2017). The creativity of literary writing. In J. Kaufman, J. Baer, & V. Glaveanu (Eds.), Cambridge handbook of creativity across domains. New York, NY: Cambridge University Press.Google Scholar
- Pelowski, M., Markey, P., Forster, M., Gernot, G., & Helmut, L. (2017). Move me, astonish me…delight my eyes and brain: The Vienna integrated model of top-down and bottom up processes in art perception and corresponding affective, evaluative, and neurophysiological correlates. Physics of Life Reviews, 21, 80–125.PubMedCrossRefGoogle Scholar
- Petty, R. E., & Wegener, D. T. (1999). The elaboration likelihood model: Current status and controversies. In S. Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology (pp. 41–72). New York: The Guilford Press.Google Scholar
- Schaeffer, J. M. (1999). Pourquoi la fiction?. Paris: Seul.Google Scholar
- Shklovsky, V. (1965). Art as technique. In L. T. Lemon & M. J. Reis (Eds.), Russian formalist criticism: Four essays (pp. 3–24). University of Nebraska Press: Lincoln.Google Scholar
- Slater, M. D., & Rouner, D. (2002). Entertaintment-education and elaboration likelihood: Understanding the processing of narrative persuasion. Communication Theory, 12, 173–191.Google Scholar
- Walton, K. (1990). Mimesis as make-believe. Harvard: Harvard University Press.Google Scholar
- Zwan, R. A. (2004). The immersed experience. Toward an embodied theory of language comprehension. In B. H. Ross (Ed.), The Psychology of Learning and Motivation (pp. 35–62). New York: Academic Press.Google Scholar