Cognitive narratology can be defined as the study of mind-relevant aspects of storytelling practices, wherever—and by whatever means—those practices occur. As this definition suggests, cognitive narratology is transmedial in scope; it encompasses the nexus of narrative and mind not just in print texts but also in face-to-face interaction, cinema, radio news broadcasts, computer-mediated virtual environments, and other storytelling media. In turn, “mind-relevance” can be studied vis-à-vis the multiple factors associated with the design and interpretation of narratives, including the story-producing activities of tellers, the processes by means of which interpreters make sense of the narrative worlds (or “storyworlds”) evoked by narrative representations or artifacts, and the cognitive states and dispositions of characters in those storyworlds. In addition, the mind-narrative nexus can be studied along two other dimensions, insofar as stories function as both (a) a target of interpretation and (b) a means for making sense of experience—a resource for structuring and comprehending the world—in their own right.
Cognitive narratology can be characterized as a subdomain within “postclassical” narratology (Herman 1999). At issue are frameworks for narrative research that build on the work of classical, structuralist narratologists but supplement that work with concepts and methods that were unavailable to story analysts such as Barthes, Genette, Greimas, and Todorov during the heyday of the structuralist revolution. In the case of developments bearing on cognitive narratology, narrative analysts have worked to enrich the original base of structuralist concepts with ideas about human intelligence either ignored by or inaccessible to the classical narratologists, thereby building new foundations for the study of cognitive processes vis-à-vis various dimensions of narrative structure.
Still an emergent trend within the broader domain of narratology (→ Narratology), cognitive narratology encompasses multiple methods of analysis and diverse narrative corpora. Relevant corpora include fictional and nonfictional print narratives; computer-mediated narratives such as hypertext fictions, e-mail novels and blogs; comics and graphic novels; cinematic narratives; storytelling in face-to-face interaction; and other instantiations of the narrative text type (→ Narration in Various Media). Meanwhile, theorists studying mind-relevant aspects of storytelling practices adopt descriptive and explanatory tools from a variety of fields—in part because of the interdisciplinary nature of research on the mind-brain itself. Source disciplines include, in addition to narratology, linguistics, computer science, philosophy, psychology, and other domains. Making matters still more complicated, because the term “cognitive narratology” is a relatively recent coinage (cf. 3), narrative scholars working on issues that fall within this domain do not necessarily identify their work as cognitive-narratological, and might even resist being aligned with the approach.
It should therefore not be surprising that, given the range of artifacts and media falling under its purview, its richly interdisciplinary heritage, and the multiplicity of projects relevant for if not directly associated with it, cognitive narratology at present constitutes more a set of loosely confederated heuristic schemes than a systematic framework for inquiry. Again, however, a trait shared by all this work is its focus on mind-relevant aspects of storytelling practices—where “mind” is shorthand for “mind-brain.” Insofar as stories constitute a target of interpretation, key questions for cognitive narratology include: What cognitive processes support narrative understanding, allowing readers, viewers, or listeners to construct mental models of the worlds evoked by stories? How do they use medium-specific cues to build on the basis of the discourse or sujet a chronology for events, or fabula (what happened when, or in what order?); a broader temporal and spatial environment for those events (when in history did these events occur, and where geographically?); an inventory of the characters involved; and a working model of what it was like for these characters to experience the more or less disruptive or non-canonical events that constitute a core feature of narrative representations (Herman 2009a: chap. 5)? Further, insofar as narrative constitutes a sense-making instrument in its own right, a way of structuring and understanding situations and events, still other questions suggest themselves for cognitive narratologists: How exactly do stories function as tools for thinking (Herman 2003)? Is it the case that, unlike other such tools (stress equations, deductive arguments, etc.), narrative is a mode of representation tailor-made for gauging the felt quality of lived experiences (Fludernik 1996; Herman 2007a, 2007b, 2009a: chap. 6)? More radically, do stories afford scaffolding for consciousness itself—in part by emulating through their temporal and perspectival configuration the nature of conscious awareness itself? In other words, are there grounds for making the strong claim that narrative not only represents what it is like for experiencing minds to live through events in storyworlds, but also constitutes a basis for having—for knowing—a mind at all, whether it is one’s own or another’s (Herman 2009a: chap. 6)?
Arguably, questions such as these could not have been formulated, let alone addressed, within classical frameworks for narrative study (but cf. Barthes 1966 and Culler 1975 for early anticipations). Cognitive narratology can thus be thought of as a problem space that opened up when earlier, structuralist models were brought into synergistic interplay with the many disciplines for which the mind-brain is a focal concern.
At the time of writing, the term cognitive narratology itself has been in use for only about a decade. As Eder (2003: 283 n.10) notes, the term appears to have been first used by Jahn (1997). (In a personal communication, Jahn confirmed that when he published this article he was not aware of any prior use of the term, but also that Ansgar Nünning must be credited with suggesting the second part of the article’s title.) However, the issues and concerns encompassed by the term have been live ones for a considerably longer period.
Beginning in the 1970s, studies in a number of fields provided, avant la lettre, important foundations for cognitive-narratological research. In the domain of literary studies, and in parallel with a broader turn toward issues of reception or reader response (Iser 1972; Jauss 1977; Tompkins 1980), research by Sternberg (1978) and Perry (1979) highlighted processing strategies (e.g. the “primacy” and “recency” effects) that arise from the situation of a given event vis-à-vis the two temporal continua of story and discourse, or fabula and sujet. Events that happen early in story-time can be encountered late in discourse-time, or vice versa, producing different reading experiences from those set into play when there is greater isomorphism between the time of the told and the time of the telling. (A still earlier precedent in this connection is Ingarden’s  account of literary texts as heteronomous vs. autonomous objects, i.e. as schematic structures the concretization of whose meaning potential requires the cognitive activity of readers.) Likewise, in the fields of cognitive psychology and Artificial Intelligence research, analysts began developing their own hypotheses about cognitive structures underlying the production and understanding of narrative.
Psychologists such as Mandler (1984), for example, postulated the existence of cognitively based story grammars or narrative rule systems. Such grammars were cast as formal representations of the cognitive mechanisms used to parse stories into sets of units (e.g. settings and episodes) and principles for sequencing and embedding those units (for a fuller discussion, cf. Herman 2002: 10–13). Roughly contemporaneously with the advent of story grammars, research in Artificial Intelligence also began to focus attention on the cognitive basis for creating and understanding stories. Schank & Abelson’s (1977) foundational work explored how stereotypical knowledge reduces the complexity and duration of many processing tasks, including the interpretation of narrative. Indeed, the concept of script, i.e. a type of knowledge representation that allows an expected sequence of events to be stored in the memory, was designed to explain how people are able to build up complex interpretations of stories on the basis of very few textual or discourse cues (→ Schemata). Whereas the term “scripts” was used to refer to kinds of world-knowledge that generate expectations about how sequences of events are supposed to unfold, “frames” referred to expectations about how domains of experience are likely to be structured at a given moment in time (Goffman 1974). Frames guide my expectations about the objects and decor that I am likely to find in a university classroom as opposed to a prison cell; scripts guide my expectations about what I can expect to happen while ordering a beer in a bar as opposed to defending a doctoral dissertation.
Although subsequent research on knowledge representations suggests the limits as well as the possibilities of the original frame and script concepts (Sternberg 2003 provides a critical review), this early work has shaped cognitive narratology from its inception, informing the study of how particular features of narrative discourse cue particular kinds of processing strategies. Indeed, Jahn’s (1997) foundational essay in the field, mentioned above, draws on Minsky’s (1975) account of frames (among other relevant research) to redescribe from a cognitive perspective key aspects of Stanzel’s (1979) theory of narrative. In Jahn’s proposal, higher-order knowledge representations or frames enable interpreters of stories to disambiguate pronominal references, decide whether a given sentence serves a descriptive or a thought-reporting function (e.g. depending on context “the train was late” might either be a thought mulled over by a character or part of the narrator’s own account of the narrated world), and, more generally, adopt a top-down as well as a bottom-up approach to narrative processing. A frame guides interpretation until such time as textual cues prompt the modification or substitution of that frame.
In a similar vein, other theorists have explored how experiential repertoires, stored in the form of scripts, enable readers or listeners of stories to “fill in the blanks” and assume that if a narrator mentions a masked character running out of a bank with a satchel of money, then that character has in all likelihood robbed the bank in question. Analysts have also discussed how literary narratives in particular involve processes of script recruitment, disruption, and refreshment (Cook 1994; Herman 2002: 85–113; Stockwell 2002: 75–89), depending on how critically and reflexively the narratives relate to prevailing scripts. For her part, Emmott (1997) focuses on how what she calls contexts, or spatiotemporal nodes inhabited by configurations of individuals and entities, constrain pronoun interpretation. Information about contexts attaches itself to mental representations that Emmott terms “contextual frames.” An action performed by (or on) a given configuration of participants is necessarily indexed to a particular context and must be viewed within that context, even if the context is never fully reactivated (after its initial mention) linguistically. For example, if a character in a short story orders a beer in a bar, then even if elements of the setting are not mentioned again readers can assume that subsequent verbal and nonverbal actions performed by the character continue to take place in the bar—until such time as linguistic signals cue a frame-switch (e.g. “Several days later, he saw his friend […],” or “Later that night, when he had reached his apartment […]”). Finally, Palmer (2004) also draws on elements of the early work on knowledge representations, studying how readers’ world- knowledge allows them to make sense of a variety of techniques for representing fictional characters’ minds. Palmer explores how readers construct inferences about fictional minds by using various textual indicators, including thought reports, speech representations, and descriptions of behaviors that span the continuum linking mental with physical actions.
More generally, a cluster of publications appeared in the second half of the 1990s, all of them adding impetus to the “cognitive turn” in narrative studies that had been prepared for by research conducted in the 1970s and 1980s and that had been directly anticipated by Turner (1991). 1996 saw the appearance of Fludernik’s richly synthetic account of natural narratology (1996), which integrates ideas from literary narratology, the history of English language and literature, research on natural-language narratives told in face-to-face communication, and cognitive linguistics to isolate “experientiality,” or the felt, subjective awareness of an experiencing mind, as a core property of narrativity (Narrativity). Turner’s (1996) own extrapolation from cognitive-linguistic models of metaphor to account for human intelligence in terms of parabolic projections, or the mapping of source stories onto target stories to make sense of the world, was also published in 1996. The year before, the influential volume Deixis in Narrative had appeared (Duchan et al. eds. 1995); contributions to this volume characterize narrative comprehension in terms of deictic shifts, whereby interpreters shift from the spatiotemporal coordinates of the here-and-now to various cognitive vantage-points that they are cued to occupy by textual signals distributed in narrative discourse (Ryan 1991; Werth 1999).
This spate of publications over a five-year period (the list is by no means exhaustive) helps explain why the inaugural 2000 issue of the online journal Image & Narrative focused on cognitive narratology. It also helps account for the organization, just after the turn of the century, of a number of edited volumes, special journal issues, and conferences exploring intersections among cognition, literature, and culture as well as cognitive approaches to narrative in particular (e.g. Abbott ed. 2001; Richardson & Steen eds. 2002; Herman ed. 2003; Richardson & Spolsky eds. 2004). During the same period, theorists formulated a number of pertinent objections to (or at least reservations about) what Richardson & Steen termed a “cognitive revolution” in the study of literature and culture (Jackson 2005; Sternberg 2003). In particular, as noted in 4 below, scholars who remain skeptical about cognitive approaches to literature and culture in general, and about cognitive narratology in particular, question the degree to which work of this kind represents true interdisciplinary convergence—as opposed to the selective (and sometimes ill-informed) borrowing of ideas and methods tailored to problem domains in other fields.
It is still too early in the development of cognitive narratology to identify what its most important contributions to the broader field of narratology may eventually prove to be. Nonetheless, the present subsection provides a partial catalogue of pertinent studies, with the following subsections focusing on several areas in which research activity has already been especially productive. Relevant research includes:
(b) research on representations of the minds of characters and on the classes of textual cues that prompt readers to draw particular kinds of inferences about the contents and dispositions of those minds (Butte 2004; Cohn 1978; Herman 2007a; Palmer 2004; Zunshine 2006);
(d) research on the range of cognitive processes that support inferences about the spatiotemporal profile of a given storyworld, and about the degree to which a given text or representation can be assimilated to the category “narrative”—that is, assigned at least some degree of narrativity—in the first place (Fludernik 1996; Gerrig 1993; Herman 2002, 2009a; Hogan 2003b: 115–39; Jahn 1997; Ryan 1991, 2003);
(e) research on the textual as well as cognitive factors underlying the key effects of narrative suspense, curiosity, and surprise, and more broadly on how the temporal order in which elements of a narrative are encountered can shape interpreters’ overall sense of a storyworld (Gerrig 1993; Perry 1979; Sternberg 1978, 1990, 1992);
(f) research more generally on phenomena pertaining to the interface between narratives and the mind-brain of the interpreter, such as the activation of “identity themes” (Holland 1975) or the (potential) stimulation of empathetic responses (Keen 2007)—in other words, attempts to formulate what Eder (2003) terms “cognitive reception theories”;
(h) empirical studies that, relying on techniques ranging from the measuring of reading times to methods of corpus analysis to the elicitation of diagrams of storyworlds, seek to establish demonstrable correlations between what Bortolussi & Dixon (2003) term “text features” and “text effects”—i.e. between textual structures and the processing strategies that they set into play (Gerrig 1993; Ryan 2003; Herman 2005); and
(i) intermedial research suggesting that narrative functions as a cognitive “macroframe” enabling interpreters to identify stories or story-like elements across any number of semiotic media (→ Mediacy and Narrative Mediation)—literary, pictorial, musical, etc. (Wolf 2003; Ryan ed. 2004; Herman 2009a).
Several of these initiatives can be singled out as especially generative for cognitive-narratological research: namely, study of the cognitive processes underlying interpreters’ ability to construct (and immerse themselves more or less fully within) storyworlds; research on issues pertaining to consciousness representation; and, relatedly, analyses of emotion and emotion discourse vis-à-vis stories and storytelling.
Mapping words onto worlds is a fundamental—perhaps the fundamental—requirement for narrative sense making. Approaches such as deictic shift theory (Duchan et al. eds. 1995) and contextual frame theory help reveal the complex cognitive processes underlying narrative ways of worldmaking; they also suggest how configuring narrative worlds entails mapping discourse cues onto the WHAT, WHERE, and WHEN factors whose interplay accounts for the ontological make-up and spatiotemporal profile of a given storyworld. An approach based on shifting deictic centers indicates how narrative worlds are structured around cognitive vantage points that may change over the course of an unfolding story. Likewise, based on the assumption that characters will be bound into and out of particular contexts over time as well as the assumption that such contexts will be distributed spatially as well as temporally, Emmott’s (1997) contextual frame theory points to the nexus of the WHAT, WHERE, and WHEN factors in narrative worldmaking.
Furthermore, reconsidered from a cognitive-narratological perspective, earlier narratological scholarship can be read anew, providing further insight into the cognitive processes underlying the (re)construction of narrative worlds. Genette’s (1972) influential account of time in narrative, for example, can be motivated as a heuristic framework for studying the WHEN component of world creation. When Genette distinguishes between simultaneous, retrospective, prospective, and “intercalated” modes of narration (as in the epistolary novel, where the act of narration postdates some events but precedes others), these narrative modes can now be interpreted in light of the different kinds of structure that they afford for worldmaking. Retrospective narration accommodates the full scope of a storyworld’s history, allowing a narrator to signal connections between earlier and later events through proleptic foreshadowings of the eventual impact of a character’s actions on his or her cohorts. Simultaneous narration, in which events are presented in tandem with the interpreter’s effort to comprehend the contours and boundaries of the narrated domain, does not allow for such anticipations-in-hindsight; rather, inferences about the impact of events on the storyworld remain tentative, probabilistic, open-ended (Margolin 1999). In short, classical, structuralist accounts like Genette’s suggest how a narrative world is “thickened” by forays backward and forward in time and throws into relief the processing strategies triggered by such temporal agglutination (Sternberg 1978, 1990, 1992).
In her foundational study of strategies for representing consciousness in narrative fiction, Cohn (1978) draws on theories of speech representation (→ Speech Representation) as the basis for her account of how narrative texts afford access to fictional minds. Just as narratives can use direct discourse, indirect discourse, and free indirect discourse to present the utterances of characters, fictional texts can use what Cohn calls quoted monologue, psycho-narration, and narrated monologue to represent the thought processes of fictional minds. Subsequent theorists, seeking to underscore even more clearly the assumed analogy between modes of speech and thought representation, have renamed Cohn’s three modes as direct thought, indirect thought, and free indirect thought, respectively (Leech & Short 1981). As Palmer (2004) notes, however, this classical or “speech category” approach captures only some of the phenomena relevant for research on narrative representations of consciousness. For Palmer, the speech-category approach has induced analysts to focus solely on inner speech, with the result that theories of consciousness representation in narrative have been “distorted by the grip of the verbal norm” (53). Yet narrative understanding in fact hinges on a wide variety of inferences about the states, dispositions, and processes of fictional minds—including inferences about the felt, subjective nature of their experience (i.e. the “qualia” specific to their particularized vantage-point on the storyworld [Nagel 1974]) as well as their folk psychology, or method for framing inferences about what is going on in their own and others’ minds.
When characters use folk-psychological models to explain their own and others’ motivations and intentions, they are drawing on fundamental, generic processes by which humans attribute mental states, properties, and dispositions both to themselves and to their social cohorts. These processes have been described as the native “Theory of Mind” in terms of which people make sense of their cohorts’ behavior (Zunshine 2006). At issue is people’s everyday understanding of how thinking works, the rough-and-ready heuristics to which they resort in thinking about thinking itself—a heuristics used to impute motives or goals to self and other and to make predictions about future reactions to events. Such thinking about thinking points beyond inner speech and solitary self-communings to the “social mind in action” that Palmer identifies as the object of study for postclassical approaches to consciousness representation (2004: 130–69).
As Stearns (1995) points out, there is a basic tension between naturalist and constructionist approaches to emotion. Naturalists argue for the existence of innate, biologically grounded emotions that are more or less uniform across cultures and subcultures (Hogan 2003a). By contrast, constructionists argue that emotions are culturally specific (Stearns 1995). As Adolphs (2005) suggests, however, the naturalist and constructionist positions can be reconciled if emotions are viewed as (a) shaped by evolutionary processes and implemented in the brain, but also (b) situated in a complex network of stimuli, behavior, and other cognitive states. Because of (b), the shared stock of emotional responses is mediated by culturally specific learning processes. Further, to study the cultural and rhetorical grounding of emotion discourse, theorists working at the intersection of psychology, history, and ethnography have developed the concept of “emotionology,” which concerns the collective emotional standards of a culture as opposed to the experience of emotion itself. The term functions in parallel with recent usages of ontology to designate a model of the entities, together with their properties and relations, that exist within a particular domain. Every culture and subculture has an emotionology, which is a framework for conceptualizing emotions, their causes, and how participants in discourse are likely to display them.
Narratives, which at once ground themselves in and help build frameworks of this sort, provide insight into a culture’s or subculture’s emotionology—and also into how members of the (sub)culture use these systems to make sense of minds. Everyday storytelling as well as literary narratives deploy and in some cases thematize emotion terms and concepts; for example, spy thrillers, and romance novels are recognizable as such because of the way they link particular kinds of emotions to recurrent narrative scenarios. What is more, stories also have the power to (re)shape emotionology itself. Narrative therapy, for instance, involves the construction of stories about the self in which the emotional charge habitually carried by particular actions or routines can be defused or at least redirected (Mills 2005).
(a) Eder (2003: 284 n. 14) sets up a scale of seven possible relationships between cognitive reception theories and narratology. These possibilities run the gamut from impossibility to unrelated coexistence to the outright assimilation of narratology to cognitive theory. A more general question can be extrapolated from Eder’s analysis: to what extent does the research conducted to date warrant commitment to the possibility of integrating narratological theory with ideas from the cognitive sciences? (b) Relatedly, Sternberg (2003) has raised questions about the degree to which cognitive narratology enables true methodological convergence among the domains of inquiry that it encompasses. Part of the problem lies in the attempt to translate foundational concepts such as “frames” and “scripts,” “emotion,” and even “narrative” across what remains for Sternberg a disciplinary divide between humanistic and social-scientific research. As this critique suggests, if cognitive narratology is to become a bonafide inter-discipline, it must work toward combining its source concepts and methods into a whole which is greater—more capable of description and explanation—than the sum of its parts.