Like most theories, narratology came under fire long before the structuralist analysis of narrative texts had been worked out in detail. The French journal Poétique, for example, was still publishing supplements to Gérard Genette’s chapters on focalization when in the United States feminist narrative theory was already in full swing. The structuralist dream of a unified, abstract, and universal theory about narrative was soon shattered by late and poststructuralist approaches drawing attention to the inevitable contradictions and self-undermining tendencies in any theory.
Contradictions have always been part and parcel of narrative theory. As Jan Baetens has demonstrated, structuralist narratology was developed in a literary and cultural context that, at least in its highbrow manifestations, was hostile toward narrative traditions.1 In France the nouveau roman famously claimed the era of storytelling had come to an end. The heyday of both structuralist narratology and experimental literature was short-lived. It lasted roughly from the mid-1960s to the mid-1970s. Then the postmodern paradigm took over, pretending that humans only have stories to come to terms with their environment and their own lives. This was quite paradoxical too. On the one hand almost everything could be considered a narrative; on the other hand the conceptualization and definition of narrative became quite hazy, as did the theoretical foundation and elaboration.
The 1970s and 1980s witnessed the so-called narrative turn.2 It manifested itself in almost all domains of the humanities, including Hayden White’s historiography and Jerome Bruner’s psycho-cultural studies.3 It led to a proliferation of narrative-inspired approaches across the board, but according to Raphaël Baroni, it reduced theory to a toolbox and subordinated it to its practical use.4 In more positive terms and following Marie-Laure Ryan, Roy Sommer says that structuralist narratology focused on syntax and semantics, whereas the newer narratologies favor pragmatics.5 While structuralists excelled in abstract theory of narrativity, newer narratologists have been more interested in concrete narrative forms and analyses. The paradox Baroni points to also has to do with the precarious institutional place of narratology, which is usually restricted to departments of literature, whereas its application is cross-disciplinary: “Paradoxically, from an institutional viewpoint, the theory of narrative appears to be dying, whereas its field of application has never seemed so wide and its utility has never seemed so evident.”6 This leads to two problems: the literary theorist does not have enough knowledge of the transdisciplinary contexts in which the theory will be used, and the scholars working in those disciplines are not well acquainted with the theoretical advances.
What is needed, says Baroni, is an integration of old and new, theory and practice, structuralist and newer narratologies. Sommer talks of a “merger” between structuralist and more recent narratologies.7 To this plea for an integrated view Baetens adds that we also need a new conceptualization and definition of narrative. As it is, narrative is usually defined in a traditional way, excluding the nouveau roman and other experiments, which tend to be relegated to the realm of exceptional (or, as we will see, “unnatural”) narratives. Moreover, those definitions are predominantly based on verbal narratives, leaving aside films, video games, and so on. When these other forms and media are studied, scholars have to transform the “toolbox,” since it was not developed for these forms and media in the first place.
Although we recognize that such an integrative approach and such a re-definition of narrative are important for the future of narrative theory, we will restrict ourselves here to a presentation of existing approaches. Moreover, as we are primarily interested in the practical relevance of these theories for the analysis of texts, we will try to restrict abstract discussions to a minimum and maximize the analytic potential of the approaches we present.
Ever since David Herman’s introduction to Narratologies (1999), it has become customary to talk of structuralist narratology as “classical” and to call the later approaches “postclassical.”8 There have been many systematizations of postclassical approaches and many characterizations of the similarities and differences between classical and postclassical methodologies, but one thing on which all of these seem to agree is the importance of the distinction between text and context.9 Classical narratology tends to limit itself to the text, whereas its postclassical successor includes the context. This context may take all kinds of forms. It may be oriented toward the agents involved in the narrative communication, thereby re-introducing the author into the field of study and giving the reader a much more important role. It may also embrace wider contexts, such as social class, gender, cultural stereotypes, and prevailing ideologies.
This attention for context need not eclipse the text. On the contrary, it may draw attention to certain texts and aspects that were hitherto left unstudied. For instance, feminist and queer narratology may address texts that highlight gender issues and may reveal those issues in texts that were thought of as being neutral in this respect. Roy Sommer helpfully distinguishes between “corpus-based” and “process-oriented” postclassical approaches.10 The former consist of, first of all, narratologies that go beyond the classical text (as, for instance, in transmedial studies) and, second, thematically oriented approaches such as feminism, as well as ethnic, cultural, and postcolonial studies. The latter include cognitive theories (focusing on the process of cognition rather than on the actual text) and rhetorical narratology (studying the process of communication between sender and receiver).
Our discussion of postclassical narratology consists of four sections. Roughly, they involve an ever-widening contextualization: the expansion moves from text to communication, to sociocultural context, and finally to everyday narration. We start from approaches that broaden the conception of narrative text. This will include intermedial studies, diachronic narratology, and theories that regard the text as a world (possible world, storyworld). In a second step we will consider the communicative broadening in cognitive narratology, which underscores the reader’s processing of narrative texts, and in rhetorical narratology, which expands the intratextual communication between narrator and narratee to the contextual communication between an (implied) author and a reader. Third, we will discuss narratologies that include the sociocultural context, such as cultural narratology and narrative ethics. In the final section, we will deal with methodologies that study our everyday lives as a narrative process. This will include a brief glance at the all-inclusive notion of “storytelling,” as well as a longer presentation of natural and unnatural narratologies.
This step-by-step discussion of postclassical narratologies will include an inevitable degree of overlap between the various sections. For example, storyworld as defined by David Herman is a concept that relies on the readerly construction of the fictional world; this means that it not only has a place in our section on narrative and worlds but could also be treated in our pages on cognitive narratology. In cases like this, we have sought to distribute information without confusing our audience.
So much is happening in contemporary narratology that it is simply impossible to discuss everything in detail. Our emphasis on the interpretive relevance of postclassical theories explains why we do not pay attention to narratological approaches that choose to discuss the context but seem to lose track of the text in the process. A similar motivation makes us somewhat skeptical toward the use of the computer in the development of the research field. Obviously these approaches have undeniable merits, but for our purposes they seem less compelling.
Anthropological narratology starts from the observation that fairy tales, legends, and myths from different cultures and periods have many characteristics in common. Following the structuralist project, this approach tends to locate the basis of this similarity in almost archetypal processes and structures such as initiation, quest, and rebirth.11 René Girard, for example, reduces stories to the triangular structure of desire (A desires an object X because the admired B desires that object) and the scapegoat mechanism (A is blamed for all social disorder and expelled so as to restore order).12 In its more contemporary form, anthropological narratology tries to explain why storytelling is such a fundamental human activity. Albrecht Koschorke argues that the benefits of stories lie in their capacity for the collective negotiation of meaning.13 Stories allow for relevant play with the distinction between truth and falsity, and they do so by “turning a state into a process.”14 The resulting sequence of events and its presentation not only manage to keep a discussion going, thanks to a story’s management of information, but they also create a form of redundancy that will, for instance, help to consolidate a (temporary) consensus. This paradox constitutes an important part of how stories may help to avoid conflict by “keeping the distance between the real and its codification in flux.”15
Psychoanalytical narratology explains stories starting from psychological processes such as displacement and condensation and from unconscious structures such as the Oedipus complex.16 The basis of a story can be found in unconscious desires that end up in the text only after various filtering processes. Examples of these include metonymical displacement (which, for instance, lifts a part from its whole and thus pushes aside those aspects of the unconscious desire that are unacceptable for consciousness) and metaphorical condensation (which, for instance, merges different people into one character).
Empirical narratology concentrates on the psychological mechanisms of text processing and almost exclusively uses positivistic methods.17 Since most narratologists have not been trained for this kind of research, the empirical study of narrative processing has largely been developed by research psychologists. However, their interest in literary narrative is fairly limited. Representative contributions to the field, such as those by Richard Gerrig and by Gordon Bower and Daniel Morrow, seem to imply that literary narratives make up too complex a phenomenon to allow for the controllable testing conditions required by the positivistic approach.18
Anthony J. Sanford and Catherine Emmott, who aim to integrate contributions on language processing and narrative comprehension in neuroscience, psychology, and psycholinguistics with work done in the humanities, are very much aware of this complexity.19 However, in their effort to create controllable empirical stimuli, they end up with a double simplification. First, they discard literary intricacy in favor of (at most) a string of relatively easy sentences that are supposed to constitute a narrative. Second, if texts go beyond such simplicity and strongly appeal to the reader for input, the density of this experience is entirely attributed to the author, who is said to program the reader for successful processing: “For communication to be effective, the writer of a narrative must ensure that the right mental models are formed in the minds of the readers at the right time. Forming the wrong one would act as a block to communication.”20 In striving for the disambiguation typical of the hard sciences, narratology might lose its relevance as a tool for the development of interpretations that ideally keep the complexity of a text intact. In any case, empirical narratology turns the narrative and literary dimensions of the text into quantitative data. Something similar happens in sociological narratology, which is based on oral stories and which connects these stories to the social group in which they originate and circulate.21 Still, this approach can be useful in literary interpretation, as we will show in our discussion of cultural narratology later in this chapter.
Contrary to anthropological narratology, psychoanalytical narratology, and empirical narratology, computational narratology does concern itself with texts rather than contexts, but it does so in a way that radically alters the premises of literary hermeneutics. While our approach in this handbook underscores the relevance of close reading for the analysis of literary narrative, computational narratology partakes of what Franco Moretti has described as “distant reading.”22 It promises to theorize aspects of narrative by analyzing big narrative corpora with the help of a computer program instead of the human brain. Next to the analysis of corpora, a broader definition of computational narratology includes “the approaches to storytelling in artificial intelligence systems and computer (and video) games” and “the automatic interpretation and generation of stories.”23
Taking his cue from “applied narratology” as it was developed by specialists of artificial intelligence, Jan Christoph Meister develops “EventParser” (a so-called “mark-up tool”), to accelerate the study of events in narrative, and “EpiTest,” for generating episodes and actions on the basis of the EventParser records.24 Path-breaking though his work may have been, Meister acknowledges that “the most important feature of [the tool he developed] is the fact that its analytic and hermeneutic operations are unable to handle any data types which have not been defined in the underlying theory and considered in the design of the program’s algorithms,” but he recuperates this apparent failure by suggesting that “any problems which occur during program development are thus particularly helpful because they identify specific gaps in our knowledge and theory.”25
As it turns out, (classical) narratology often provides the basis for the kinds of ambitious inquiry that are under way in this rapidly growing field. Annelen Brunner has attempted to devise a program for the automatic recognition of “speech report.”26 In a first step, she devises for its various types (including free indirect speech) an annotation system that rests entirely on narratological research and does not attempt to tackle its unresolved questions. As Brunner is wise to mention, the decisions required of the “human annotator” can therefore be “fraught with uncertainties.”27 Free indirect speech in particular is hard for a computer to recognize because it does not have any “indicators that are frequent, stable and clear enough.”28
In postclassical narratology, the narrative text has received an expanded meaning. It now includes narratives in various media (see section 1.1), its historical dimensions now come to the fore (section 1.2), and it has assumed the properties of a world (section 1.3).
In the early days of structuralism, narratologists looked for universal structures that were supposed to be at the root of all narratives, irrespective of their concrete medial forms. So, in principle, classical narratology was meant to be applicable to all kinds of narrative forms and media, including film, photography, and fashion. In practice, however, most classical contributions focused on verbal and textual narratives. This was not surprising, given the linguistic, Saussurean foundation of classical narratology and the widespread idea that the human species had become the “storytelling animal” thanks to language.29 Especially when narrative is defined in a strict sense—calling for setting, characters, and cause-and-effect-relations—language is the preferred medium of expression. Marie-Laure Ryan talks about “the primacy of language as narrative medium.”30 French scholars like Roland Barthes, Gilles Deleuze, and Jacques Derrida in their later, poststructuralist studies did focus on the narrative power of nonlinguistic media, on photography, film, and drawings on a postcard, respectively. But this was not a systematic elaboration of classical narratology and seemed to belong to an altogether different framework. By that time, classical narratology had embraced the primacy of verbal narrative thanks to the founding work of Gérard Genette, who in the early 1970s built his narratology on the study of Marcel Proust’s fiction.
However, that same period saw the beginning of the “intermedial turn.”31 Both in the cultural and the theoretical field, the role of media and their manifold interactions became central. The early, politically inspired media studies of Marshall McLuhan and Walter Ong were soon followed by a host of studies, from all disciplines, focusing on dance, opera, graphic novels, video games, and other arts.32 In this handbook it is impossible to cover all the arts and the disciplines that study their narrative aspects. In order to cover the changes and additions to classical narratology that the focus on different media may entail, we have decided to concentrate on comics.33 The plethora of theories has also led to an immense collection of new terms, which have been used in a lot of different and not always compatible ways. For our purpose, it will suffice to clarify the central terms that contribute to our narratological analysis.
A medium is more than a neutral channel that passes on information in communication. It forms and transforms the content and the communication process.34 As such, it is not a one-dimensional entity. Siegfried Schmidt distinguishes four dimensions in a medium: the semiotic system it makes use of (e.g., music or language), the technological aspect (e.g., digital technology for video games), the socio-institutional dimension (e.g., a newspaper for serial comics), and finally the relation of these first three dimensions with the other media available at the time.35 Following Ryan, we will limit these four dimensions to three—the semiotic, the material or technological, and the cultural.36
Intermediality in the widest sense of the word “designates those configurations which have to do with a crossing of borders between media.”37 This crossing can take place within one work or between various works. Let us first look at the internal crossings. A novel that contains drawings and pictures would be labeled “plurimedial” and “intracompositional” by Werner Wolf.38 However, to avoid talking about plurimediality within one medium, we will adopt the term “multimodal” novel, suggested by Günther Kress and Theo van Leeuwen.39 This may raise the thorny topic of the relation between medium, modus, and genre, but we will use the term “modal” only to refer to a medium within a medium.40 For instance, photography and drawing contained within one work of fiction are two modes. It might be argued that most media are multimodal. Ruth Page even proposes “to reconceptualize all narrative communication as multimodal.”41 Multimodality is obvious in the case of opera (combining text, music, and drama) and comics (combining linguistic and visual media), but it is also part of oral storytelling (voice and gestures) and of poetry (sound and language). If one includes the material aspect of novels, such as the cover design and illustrations, every novel could be labeled multimodal.
A second case of intermediality within one work is “intermedial reference,” or the medial form of intertextuality.42 In this case one medium refers to another one, implicitly or explicitly, on the level of content or form, but this does not result in a multimedial (i.e., “extracompositional”) work of art. A narrative text can include overt or covert references to music and films; it may even imitate the formal characteristics of these media, for instance in novels that imitate montage structures, or in visual poetry, where the text becomes a sort of painting. Wasco’s “City” contains drawings of works of art, exhibited in the streets. These can be regarded as explicit intermedial references on the level of content. They are like quotations.
Second, medial interactions exist between different works of art. The clearest form of this is “intermedial transposition,” in which a source medium is transformed in and through a target medium.43 Examples are the novelization of a TV series or, vice versa, the adaptation of a novel for the screen. More generally, media may be mixed in performances and exhibitions, forming the “extracompositional” equivalent of intracompositional multimodality.44 At the outer margin of intermediality one might situate “transmediality.” That term covers all “phenomena that are non-specific to individual media.”45 Again, these may appear on the level of content (for instance, themes and characters can be found in all medial kinds of narrative) and of form, such as in metalepsis, which can be found in fiction, film, and painting. One could say that the universalistic aspirations of early structuralist narratology were aimed at these transmedial dimensions, since it wanted to uncover narrative dimensions that supposedly appeared across all media.
The nature of media precludes universal and essentialist definitions and characterizations.46 A medium is never simply an abstract semiotic system, because it implies a certain technology, materiality, and (institutionalized) context. As technologies and contexts change, media change. Photography of the predigital age is not the same as its digital version; its social and cultural function is completely different too. It is no coincidence that media studies came to the fore at a time when the prestige of “elitist,” highbrow art was dwindling. Media that used to be regarded as lowbrow, in the sense of commercial, popular, and simplistic—one might think here of TV soap operas or sentimental photo novels—started to receive serious attention from academically institutionalized disciplines and scholars. Ryan regards McLuhan and more generally the cultural context of the 1960s as “instrumental in breaking down the barrier between elite and popular culture, a move which led to the emancipation of media studies from literature, philosophy, and poetics.”47 It is of course a more complex phenomenon, one in which media and context influence one another. As Van Leavenworth observes, “Cultural studies, media studies, and several sociopolitical critical perspectives have led to a revaluing of popular culture artifacts.”48
The multidimensional, fluid, and constantly changing nature of media poses serious challenges to narratology, most obviously to the classical toolkit, which, as we indicated, was designed with abstract, unchanging, and transmedial concepts in mind. To assess the usability of narratological terms for the study of media and intermediality, Marie-Laure Ryan and Jan-Noël Thon propose a continuum. On the left side they situate notions that are “medium free” (or transmedial), such as setting. On the right side one can find “medium-specific concepts,” such as the gutter, referring to the empty space between panels in comics. In between there are concepts that are “transmedially valid yet not medium-free.” Those need extensive revision if they are to be transferred from one domain (say, classical narrative studies of fiction) to another (say, film studies). Ryan and Thon mention “interactivity” as an example, which is different in all media and seemingly absent from others (such as traditional fiction).49
When one transfers concepts from one discipline to another, they often end up being used in a metaphorical sense. The plot of a symphonic poem is not really a plot like in a detective novel. One can study musical progressions, themes, harmony, and counterpoint as if they all contributed to the construction and development of a plot, but the important words here are “as if.” Studies in musical narrative abound with metaphorical combinations of narratological and musicological terms and ideas. Eero Tarasti, for instance, uses A. J. Greimas’s semiotic square and Wayne Booth’s implied author to turn music into a narrativized communication.50 Tonal progression and plot structure are linked, such that modulations look (or sound) like narrative twists.
This metaphorical nature is redoubled when the original concept is metaphorical right from the start. The narrator is a good example. There is no literal or real narrator in a novel, but it makes sense to study the narrative as if it were being spoken by a narrator. If this concept is transferred to film or comics studies, its “as if” qualities become even more apparent.51 With intermediality the metaphorical character is redoubled once more. If dance gestures try to imitate stills of a film or if they look like paintings, this reference is to be taken not literally but metaphorically.52 Likewise an experimental novel combining collage and montage transforms these techniques associated with painting (or photography) and film, respectively. It may remind the reader of paintings and films, but it is of course not the same.
The reader, largely absent from classical narratology, receives a lot of attention in media studies—again in a not-quite-literal way. The reader is often replaced with a viewer, a listener, and, in the case of digital fiction and video games, an interactive agent, a player, or a conarrator or coauthor. It is hardly surprising then that media studies go well with the postclassical attention for the active role of the so-called receiving agent in narrative communication.53 A clear example of this active role is to be found in computer games, studied via cybernarratology.
Cybernarratology is mainly concerned with so-called “hypertexts,” that is, all kinds of digital texts that collect data in a network in which a (potentially infinite) number of nodes are connected to each other in a (potentially infinite) number of ways. Apart from language, such elements as graphics, sound, and video material can be part of the hypertext. Well-known examples are video and computer games, multimedia stories, interactive texts, and websites. Espen J. Aarseth created the neologism “ergodic literature” for these types of text, “using a term appropriated from physics that derives from the Greek words ergon and hodos, meaning ‘work’ and ‘path.’ In ergodic literature, nontrivial effort is required to allow the reader to traverse the text.”54 The delineation and definition of the corpus is open to discussion, but all studies agree on the two characteristics we briefly discuss here.55
First, different layers of the cybertext are often visible at the same time, for instance when a mouse click conjures up another text. This can be related to the postmodern notion of the text as a palimpsest. Palimpsests are pieces of parchment that bear traces of texts that have been effaced. When a new text is written on the parchment, the earlier texts are still evident. Even though this image has been canonized by the structuralist theoretician Genette, it is especially in postmodern literary theory that it has become a popular notion, because this paradigm assumes that every text rewrites or overwrites other texts.56 Small wonder that hypertext was welcomed as the palpable and concrete fulfillment of postmodern ideals such as network-like intertextuality and the endless production of meaning.
Especially in the beginning of the 1990s, hypertext prophets such as George Landow and Jaron Lanier stirred up a nearly euphoric mood.57 It almost seemed as if the new text type constituted the beginning of a total liberation not only from the constraints of paper text but also from social reality. In this connection Lanier’s notion of “virtual reality” is very important.58 Hypertexts were claimed to present a different kind of reality in which things are realized that are merely possible in the real world—if they are not improbable or even outright impossible. These prophecies never came to much, and the notion of “virtual reality” lost much of its appeal in the decade after its introduction. In the preface to a revised edition of Narrative as Virtual Reality, Ryan declares that the update was necessary because of some important developments in the previous ten years: “The most important of these changes is the disappearance of virtual reality (VR) from the radar of the media.”59
In some cases cybernarratology limits itself to the design of new terms and metaphors that give narratological discourse a fancy touch but that do not really contribute to the theory.60 According to Aarseth, this type of cybernarratology all too often boils down to a terminological trade-off in which cyberterminology is imported into literary theory and terms from literary theory are exported to the study of cybertexts.61 The difference between the textual and the hypertextual world is ignored, even though the dimensions of time and space, for instance, are clearly different in the two worlds. Hypertexts showcase a visual and in certain applications even a tangible world representing time and space concretely, which is not the case in literary works. In fact, in the literary text, time and space are no more than metaphors, while traditional narratology pretends they are real—as if these texts actually staged a time, a space, and a world. Aarseth aims to correct these metaphors in literary theory, criticizing them from the perspective of hypertext studies.62 This goes well beyond the familiar criticism of the structuralists’ spatial three-layered model because Aarseth questions the world as it is construed by structuralist narratology at the level of the fabula. In order to resolve this problem, Aarseth develops a pragmatic model in which texts are no longer conceived of as worlds but as communication processes.
This brings us to the second crucial characteristic of hypertexts: the importance of the reader, who often becomes a player. In most cases this importance is theorized by means of the concepts of immersion and interactivity. Precisely because of their active involvement, reader-players lose themselves in the computer game they are playing or in the digital text they are writing with the help of all kinds of computer techniques. According to Ryan, this combination of immersion and interaction is not possible with literary texts. Literary texts that force the reader to participate actively—textes scriptibles or “writerly” texts, to use Barthes’s terms—inevitably shatter the effects of realism experienced by readers: they introduce distance and lead readers to consider literary procedures more closely, which disrupts the immersion.63
Ryan relates immersion to the phenomenological approach of reading as a complete conflation of subject (reader) and object (text). She connects interaction with the structuralist approach of the text as a game, a system of rules that induces action. As a combination of immersion and interaction, hypertext would be an object of investigation in which the two traditionally opposed approaches could meet. This would imply a reconciliation of the phenomenological conception of the text-as-world with the structuralist view of the text-as-game. Ryan starts from this perspective on hypertext to enrich literary narratology. She is looking for narrative strategies that are geared toward immersion or tries to find strategies that aim to achieve precisely the opposite effect. She also sheds light on the paradoxical attempts to create the illusion of a hypertext in a text—a short-term illusion of the synthesis of reflection and immersion. In this way cybernarratology increases our understanding of the literary communication and reading process.64
Seen from this perspective, hypertexts demonstrate what literary texts do to a reader in an extreme and paradoxical way. In her now-classic study Hamlet on the Holodeck, Janet Murray reads the digital narrative text as an extreme version of the stories readers were confronted with before the digital “revolution.” The immersion in a strange world as well as the possibility of interaction are much more manifest in digital text types than in nondigital ones. Murray relates this to an additional characteristic of hypertexts: the ease with which the fictional world can be adapted.65
The idea that cybertext may be linked to extreme forms of traditional texts is also present in Markku Eskelinen’s Cybertext Poetics, probably the most elaborate attempt to adapt narratology to the needs of ergodic literature. Eskelinen complains that narratology, even in its postclassical guise, developed its insights by disregarding experimental forms of storytelling and by reducing narratives to simplified forms of communication. His alternative narratology starts from Genette and from playful literature found in postmodernism (which, as we mentioned, exploits the palimpsest nature of the text) and in the French OuLiPo (Ouvroir de Littérature Potentielle, or Workshop for Potential Literature) experiments in a form of constrained writing that turns the text into a play with scripts and rules, very much like the ludic dimension found in hypertext.
In a first step this leads to an adaptation of Genette’s concepts, especially on the level of time, focalization, speech representation, and narration. In a second step Eskelinen proposes a new kind of ludology, one that makes use of some of his adaptations of Genettean concepts but is much broader than any approach focusing on the analysis of narrative. This is not surprising: Eskelinen does not believe ergodic forms like video games can be studied fruitfully by looking at them as narratives. They are much more, and sometimes they are completely different: “games are not narratives or stories (although they may include, adapt and embed elements of them).”66 There is a wide spectrum of views on the proximity of narrative and narratology on the one hand and computer games and ludology on the other. These range from “pretty close” to “far apart.”67 Since we want to remain close to our narratological base and purpose, we restrict ourselves to the first aspect of Eskelinen’s approach.
To Genette’s idea of fictional time (“pseudo time”) as a relation between narrated and narrative time, Eskelinen adds a second layer of “true time,” consisting of system time and reading time, both of which can be ascertained empirically in the case of cybertext: “system time to account for the varying degrees of the text’s permanence (the appearances, disappearances, and possible reappearances of the text and its parts and phases), and reading time for the text’s potentially limited and controlled temporal availability and accessibility to the reader.”68 In addition, Eskelinen posits a third layer, “real time,” which he defines as “the real-time textual communication among users within the text as an essential part of its consumption and construction.”69 These new temporal dimensions influence (the relation between) narrated and narrative time and therefore affect order (Eskelinen lists no fewer than twelve types of order), frequency, and duration.
To deal with speech in cybertext, Eskelinen includes “user’s discourse” in addition to “narrator’s discourse” and “character’s discourse.”70 Focalization too would have to include the actions of the reader-user, as these influence the access he or she gets to the character’s interior.71 Because of the multimodality of cybertexts—which may combine auditive, visual, and written signs—the narratorial system of Genette has to be broadened to include “a new category of bidiegetic narrator for narrators that either reversibly or irreversibly shift their position between homodiegetic and heterodiegetic positions.”72 Genette’s system must also become more flexible, accommodating the various types of dynamic change that occur in narrating agents.73 But even after all these transformations, Eskelinen states that narratology needs to be complemented and in some cases supplemented by ludology.74 Oxenfree, for instance, has dialogue bubbles appearing over the central character’s head, thus giving the player a choice of two or three dialogue options—a narrative situation narratology is not able to address.
Comics, or drawn strips, as the literal translation of the French term bandes dessinées would have it, are another example of a multimodal medium (combining visual and linguistic sign systems) that has received a lot of attention since the late 1970s and increasingly since the 1990s, both from narratology and from other disciplines.75 The research field has become so wide and complex as to preclude any form of representative discussion in the context of our handbook. We will restrict ourselves to some notable theories that are linked to the narratological enterprise central to this handbook, ranging from semiotics (Thierry Groensteen) and Genette-inspired views (Kai Mikkonen) to cognitive studies (Karin Kukkonen). As we go along, we will point to other approaches, but they will remain in the background.
All of these studies recognize the crucial role of the reader, as the art of comics is an art of gaps, of things not mentioned and presented. Jared Gardner expresses a general attitude: “Of all narrative forms, comics are in many respects the most inefficient, a form that depends as much on what is left out as on what is included—and a form that depends on an active and imaginative reader capable of filling in the gaps in time.”76 With the reader, the context enters the research field. However, since we focus on narrative analysis, we will not deal with the wider social, cultural context, and history of comics. Neither will we discuss its material and technological side. Of Ryan’s three media dimensions (semiotic system, technology, context), we will zoom in on the first, though the context becomes more important in the cognitive theory we present later.77
Prototypical comics are narrative in nature, so the link with narratology is not far-fetched.78 In the conclusion to his classic The System of Comics (2007), Thierry Groensteen complains that “narratology suffers from having developed in reference only to literature, when its field of natural investigation is in reality the narrative genre, and should no longer exclude the art of visual stories.” He adds that narratology “should tackle narration in confronting the entirety of disciplines of the story. The widening of research into comics (and into the photo-novel) can only lead to the necessity of modifying or revitalizing certain concepts.”79 As a result, there can be no universal theory of narrative applicable to all media. Groensteen sums this up when he discusses the necessity of adapting the concept of narrator in its application to the study of comics: “I do not believe in the possibility of establishing a general science of narratology that would be valid across all types of narratives in whatever medium.”80
Groensteen’s modification of concepts is not inspired by Genettean narratology but by a macro view on semiotics. Groensteen labels his approach “neo-semiotic.”81 He does not share the semiotic concern with defining ever smaller basic units but instead focuses on the systematic relations between pragmatically defined units. His is an “approach from on high, from the level of grand articulations.”82 As Hannah Miodrag shows, this break with structuralist semiotics and linguistics follows naturally from the visual sign system, which is not, like language, reducible to minimal units. Images are not arbitrary (since there is a link between signified and signifier, the drawing and what is shown by it), not binary (as in minimal pairs in language, e.g., voiced versus voiceless), and not discrete (there is no discontinuous minimal unit in an image or a drawing). Visual systems do not function like linguistic ones, as there is no abstract langue that is turned into concrete parole.83
In terms of Ryan’s three media dimensions (semiotic system, technology, context), Groensteen deals only with the first one: “Comics will be considered here as a language, that is to say, not as a historical, sociological, or economic phenomena [sic], which it is also, but as an original ensemble of productive mechanisms of meaning.”84 This original ensemble consists of units and relations. The basic mode of the medium is visual, not linguistic. Groensteen insists that the image is more important than the word in comics.85 It certainly is in our Wasco example, since that contains no words apart from the title. As “narrative drawing,”86 the art of comics is a “predominantly visual narrative form,”87 with “the image as a base unit.”88 Groensteen writes that “indeed, compared to a literary story, the image translates and expresses in visual terms all that it can: characters, décor, objects, atmospheric notations, expressions, gestures, actions—everything, in reality, except verbal exchanges (and thoughts), which the image is not able [to] translate and can do nothing but cite.”89
In comics the basic form of the image is the panel. In Wasco’s “City” every tier (or strip, or row) contains four panels, apart from the title and the final tier, which both consist of just one panel. A panel is delineated by a frame and separated from other panels by a white space between the frames (the gutter). In Wasco’s work, however, there is no white space: the panels are linked by the vertical line of the frame, and they seem cramped and pushed together. This suggests an intense contact between the panels, which is belied by the often disparate scenes and points of view to be found in the panels. For instance, the first tier shows four different locations and four different perspectives. This combination of continuity and disparity is typical of comics, which, as Gardner says, are “always rooted in the narrative structure of shocks, fragments, and discontinuities.”90 In Wasco’s case the combined continuity and disparity may further highlight the paradoxical nature of the city: every sight is different, yet they all belong to the same site.
The part of the page that is used by the panels is delineated by the hyperframe, which separates the panels from the margin, that is, the remainder of the page, typically (but not necessarily) left blank. With Wasco, the hyperframe is clearly drawn in black lines enclosing the title panel, the twenty panels depicting the visit of the main character and the final panel showing the skyline of the city. A set of panels makes up a “multiframe.”91 With Wasco, one can regard the complete drawing as the overarching multiframe and the twenty panels of the visit as an embedded multiframe, showing a visit to a city. We will refer to that central part as the “visit multiframe.”
In between the visit multiframe and the hyperframe, there is a small strip filled with thirty drawings of the main character and twenty-eight of his or her dog. The smallness of the strip and the large number of replicated figures may be indicative of the crowded and claustrophobic nature of the city, while at the same time it may suggest that there are really only two living beings in this environment. It is as if two prisoners are allowed to walk on a very small strip in a confined area. At any rate, the frames and hyperframe suggest a doubly closed space, with no exits. Interestingly the first and last panel of the visit show that the main character needs a spacecraft to enter and leave the city. The geometric pattern of the visit multiframe, containing twenty panels of the same size, enhances this impression of a closed space. The multiframe looks like a perfect table with four columns and five rows. The complete absence of language may be read as sign of estrangement and lack of both humanity (there are no humans apart from the main character) and communication.
Groensteen compares the spatial division of the multiframe to a beat becoming faster, with more panels per line.92 With Wasco we have a regular beat, which in terms of musical time signature could be read as four/four: there are four beats in a measure and every beat consists of a quarter note. That is the so-called common measure, used in most pop and rock music, which again may underscore the machine-like order of the city visited by the protagonist. It comes close to what Groensteen calls “the cadence of the waffle-iron grid.”93 To measure that cadence more precisely, one could look at the amount of time presented by the panels. For instance, three panels might cover only a moment (e.g., someone getting on a bus) or perhaps a long period (e.g., the change of seasons in a year). This is comparable to the classical distinction between narrated time and time of narration, except that the visual presentation here replaces the linguistic one.94 Eric S. Rabkin tries to find equivalents to the literary modes of description (slow), dramatization (even), and summary (fast) by looking at complexity (slow), representation (even), and montage or symbolism (fast). In Wasco’s “City” the last option seems to prevail: we get a montage of scenes, leaving out a lot of things, and as a result the pace seems to be quite fast.95
With the multiframe we have moved on from the level of basic units to the systematic level of their relations. Put abstractly, all images are held together by their “iconic solidarity.”96 They are at the same time separate and similar; they are placed next to one another but they remain distinct, multiple. Phrased more concretely, the relations between panels are “spatio-topical”; they all have their own place (“topical”) in the space (“spatio”) constructed by the multiframe. This positioning can be characterized in terms of form, area, and site.97 With Wasco, the form of each panel in the multiframe depicting the visit is square—again underlining the repetitiveness and the closedness of the city space. There are no variations in the panel forms, suggesting they are interchangeable, impersonal, indistinctive—characteristics that might easily be transferred onto the city. The area covered by each panel is interchangeable, as they are all of equal size. The site is, of course, different for every panel, as it refers to the unique position of the panel in the multiframe.
To uncover the spatio-topical relations between panels, one must not only look at the contiguous relations between, for instance, panels in one and the same tier, but one must also pay attention to the relations between distant panels. A panel in the middle of the story may be an echo of a panel at the outset of the narrative. Groensteen calls this articulation of the manifold links between panels “arthrology.”98 The contiguous relations are studied under the heading of “restricted” or “restrained” arthrology. The panel is articulated into a syntagm (a combination of the panel with the preceding and the following ones) and further on into a sequence (a larger narrative unit, not unlike Barthes’s functions, depicting, for instance, a complete action).99
In the case of Wasco the sequence consists of a visit to a city, and the syntagms range from abruptly changing scenes (as in panels one, two, and three) to more fluent transitions that suggest a progression in movement (as in panels nine, ten, and eleven). The two dimensions may be integrated in a reading that calls the movement in this comic abrupt and unpredictable. Although the whole sequence clearly describes a trip through a city, the progression and order of the smaller scenes or panels seem hard to determine, as there is often no discernible path or direction, no compelling or obligatory order. For instance, the order of panels fourteen to nineteen could easily be changed without anyone noticing an anomaly. It seems that the city also affects the sense of teleology and direction.
As it articulates the similarities and differences between panels, arthrology is always concerned with movement and dynamics. This movement is not (only) linear; it goes back and forth between panels. The first panel may receive its meaning only after the reader has looked at the second and the third. Groensteen talks about “retroactive determination” in this context.100 Many recent narratological analyses of comics, including those by Miodrag, Mikkonen, and Kukkonen, have stressed that reading comics is not necessarily a sequential process and that the production of meaning in comics should not be reduced to an unfolding of elements in a sequence. The eye of the reader sees different panels at the same time; the reading order is often not horizontal and sequential. For instance, in the case of inserts, vertically ordered panels or two horizontal tiers that should be “read” at the same time as the lower one show what is left out in the higher one. Following the French semiotician Pierre Fresnault-Deruelle, Charles Hatfield talks of “tabular reading” as opposed to linear reading.101
Comics may flaunt spatiotemporal rules by showing different periods in one and the same panel or by making it impossible to link the spatial setting with any clear indication of period and time. As Miodrag says, “the sheer variety of different sorts of networked relationships” in comics “casts considerable doubt on the ‘space equals time’ epithet” that is essential to sequence, where progress in space equals progress in time.102 As we said, the progress in Wasco’s “City” is not always clear, though the general movement and space are clear: we are witnessing a touristic walk through a city. The time during which this visit takes place is probably somewhere in the future, but there is no way of dating it exactly. Nor is it possible to determine the duration of the visit or of the scenes shown in the various panels. The overarching spatial and temporal categories may be relatively clear (“city” for space and “sometime in the future” for time); the specifics of the two, as well as the relations between them, are vague.
Restrained arthrology focuses on horizontal and contiguous relations, privileging traditional narrativity. General arthrology, on the other hand, focuses on distant relations and places panels in a tabular relation to one another. Groensteen calls this articulation “braiding.”103 In it, panels that are far apart are linked to form a network in a way reminiscent of weaving threads into fabric. Braiding combines the horizontal with the vertical, the synchronic (the co-presence of the panels that echo each other) with the diachronic (the evolution between the two panels). In Wasco’s “City” the last panel of the visit echoes the first via the spacecraft. It rounds out the narrative, making it almost circular—which the reader may see as yet another confirmation of the closed space suggested by the squares. The craft is absent from all other panels apart from the final one, which does not belong to the multiframe of the visit. The title panel evokes all the panels with the main character and the dog, but contrary to all of these, the two characters are separated (by a road) in the title panel. Maybe this underscores the lack of contact in the city.
All comics studies agree that the Genettean concepts of narrator and focalization cannot simply be transferred to the study of comics. The basic reason is the twofold nature of comics: they combine word and image, whereas narrator and focalization are designed to keep these separate. Telling and showing are combined, but classical narratology tries to keep these two apart.104 A narrator in Genette’s sense cannot account for the visual or pictorial track of the comic, and a focalizing agent cannot account for the various voices. As Achim Hescher says, “if the verbal and the pictorial track are considered as one [ . . . ], ‘focalization’ can neither account for one nor for the other because no fictional entity from the mediating/transmitting communication system generates both tracks, that is[,] the narratorial script as well as the images (including the speech/thought balloons).”105
Many different adaptations of the narrator/focalization terminology have been proposed. We have already referenced Groensteen’s “fundamental narrator,” embracing the “monstrator” and the “reciter.” Hescher “rejects the use of the concepts ‘narrator’ and ‘focalization’ from classical narratology.”106 As to focalization, he takes his cue from François Jost’s “ocularization.”107 This is a term indicating the relation between what is shown to the reader in an image and what the character actually sees. More specifically, he distinguishes between “primary internal ocularization,” representing what a character sees without any sign of that character; “secondary internal ocularization,” where we get a glimpse of the person looking; and finally “zero ocularization,” where there is no subject attached to the image.108 In “City” the panels usually show the main character, which is of course not what that character actually sees (he or she would need a mirror to see himself or herself). These would be examples of zero ocularization. Only panels two, three, and nineteen may be rendered through the eyes of the main characters, in the form of primary internal ocularization.
As to narration, comics tend to be multivoiced. Complementing Ann Miller, Hescher takes into account six types of texts in comics: paratext, narratorial captions, balloons and balloon speech, lettering (as in onomatopoeic signs like “boohoo”), intradiegetic texts (e.g., a letter shown in a panel), and nondiegetic tags, or seemingly free-floating comments in an image.109 The latter usually originate from “the artist-writer,” as Hescher calls the ultimate narrating agent responsible for both the verbal and the visual track.110 This complexity of the “verbal-narratorial track” cannot be subsumed under the heading of a narrator.111 At any rate, the words in a graphic text are always visual elements too, and a classical narrator cannot deal with that.
Kai Mikkonen seems to agree with Hescher. He says it makes no sense to create “a separate agent” for “narration through showing,” as opposed to an agent for narration through telling. The two operate in combination (which may take many forms, ranging from juxtaposition to reinforcement) and in the end derive from “the author as narrator.”112 Mikkonen’s The Narratology of Comic Art may be one of the most extensive dialogues between the study of comics and classical, Genettean narratology. He deals with all the aspects we have covered in chapter 2, starting with the temporal organization (where he uses the relation between narrated time and narrating time to decide temporal order, duration, and frequency), moving on to characterization, focalization, speech and thought representation (with an interesting take on the visual means for suggesting the simultaneity of two perspectives present in free indirect speech), and narration. However, in all of these elements he changes, adapts, and sometimes supplants the structuralist concepts to adequately study the mix of words and images that distinguishes comics. For instance, dialogue is more than words in comics; it is also facial expressions, gestures, symbolic signs (such as the “emanata” around the head of a character)—in short, it is embodied.113 Moreover, Mikkonen also considers aspects that are not dealt with in classical narratology, such as (graphic) style and spatial anchorage of perspective.
Spatial positioning is much more prominent in a visual medium than in a verbal one. Any image in a comic is always seen and presented from a certain angle, which is not necessarily the case with verbal descriptions.114 In Wasco’s “City” the view we get of the main character is usually from a high and quite distant vantage point. If he or she were filmed with a camera from that position, it might very well resemble the kind of surveillance camera we find in every city. We never get any shots from the same height as the character, we never get any close-ups, and we cannot even see the face of the character or the distinguishing features of the dog. All of that creates a distance between the reader and the character. There is no “proximity,” no “alignment,” which, according to the definition of Murray Smith, describes “the process by which spectators are placed in relation to characters in terms of access to their actions, and to what they know and feel.”115
In fact the characters are not really individualized; they look like caricatures or types—which may hark back to the historical link between comics, cartoons, and caricature. Moreover, as a result of the distant and high camera view, the characters are invariably small. All of this may underline the fact that the city has become an anonymous world, reducing humans and animals to insignificant and impersonal beings. Mikkonen discusses a wide array of “mimetic images”—visual means of expressing the inner world or the mind of the character.116 Almost none of those are present in Wasco’s narrative. The visitor behaves like a typical tourist: pointing to interesting sights (panel six), walking about endlessly, relaxing only once (panel seventeen). Whether the visitor finds this glorious, boring, or appalling, readers have no way of telling, unless they take the rather gloomy exterior world of the city as a reflection of the character’s mind. But that would be the reader’s responsibility.
A step further removed from classical narratology and many steps closer to the reader is the cognitive approach to comics, exemplified by Karin Kukkonen’s Contemporary Comics Storytelling. As she states at the outset, “this marks a shift in perspective from the code-based comics semiotics of Thierry Groensteen and others.”117 Comics are no longer studied as a code for specialists but as a form of meaning production that involves three parties: the graphic text, the reader, and the context. Taking her cue from the basic characteristic of comics as an art of gaps, Kukkonen claims that the graphic text is a set of “clues and gaps” that incite the reader to activate certain cognitive schemata, which tie in “with basic cognitive processes” we use in everyday life.118 These basic processes are both conceptual and preconceptual, involving knowledge (e.g., of genre, people, the social world) and experience (corporeal, embodiment, and so on). They mediate between text, reader, and context.
As we will see when we discuss the cognitive approach in more detail, this mediation is seen as a double process that combines text-induced inferences (bottom-up) with the reader’s activation of cognitive schemes (top-down). Kukkonen clarifies this process through an analysis of three comics series. She studies them as (re)negotiations of postmodern topics in present-day society via the activation of specific schemata. Fables conjures up the generic frame of fairy tales, which are drastically rewritten (e.g., turning Prince Charming into “an opportunist and exploiter of his own myth”) and combined with incompatible genres.119 The series shows the importance of intertextuality in the processing of comics and raises the issue of good and evil in postmodern, relativist times. Tom Strong, the second case study, activates the schemata of the superhero and does not satisfy itself with the cliché that such heroes are simplistic figments of the imagination. It raises the question of fictionality and reality in the postmodern, panfictional world. Finally, 100 Bullets incites readers to enter the minds of the characters, especially of murderers-out-of-revenge. It “can be read as a negotiation of ethics in the postmodern world,” a world that has supposedly witnessed the “end of ethics.”120
Kukkonen focuses on concrete visual and textual forms—such as close-ups, drawing style, and coloring—not because she wants to lay bare a semiotic code or present a grammar of concepts but because she wants to link these graphic clues and gaps to the reader’s schemata that relate the text to the context. Typically, this link is metaphorical. For instance, “three ‘conventional metaphorical mappings’ can be discerned as particularly prevalent in 100 Bullets: social life is an economic transaction, social life is a game, and social life is war.”121 As a result of this focus on readerly processing, classical concepts such as narrator and focalization recede into the background. Kukkonen alleges that “narrators and focalization are generally not as easily identifiable in comics.”122 But even though she claims that “there is rarely a discernible narrator in comics,” in her conclusion she admits that a narratology of comics should also deal with the issue of narrators and focalization.123
Kukkonen’s detailed analysis may activate schemata in Wasco’s readers. Among other schemes, intertextual and generic knowledge seem to play an important role here. The title panel, showing the main character and his or her dog, may remind readers of one of the iconic couples in the history of comics—Tintin and Snowy. This reference may become more compelling when one reads the remainder of the album Het Tuitel complex. Near the end of the book there is a six-page album that explicitly refers to Hergé’s characters and stories.124 The cover of this album-within-an-album is a transformation of Hergé’s cover, with the name of that author reduced to “H.” The protagonist’s hair style is changed and becomes the same as Tintin’s. The title of the album is “America,” which refers not only to Hergé’s classic Tintin in America but also to Kafka’s Amerika. The story itself consists of drawings without texts; every page contains nine multiframes (each looking like smaller pages) that abound with allusions to scenes from both Hergé and Kafka. The final page of Wasco’s album is a rewriting/redrawing of the typical Hergé back cover, combining Hergé with Kafka (in the titles quoted, e.g., The Metamorphosis), as well as with the popular Flemish comics author Willy Vandersteen, whose Bob and Bobette are present, together with Tintin and Captain Haddock.
The evocation of Tintin in the “City” narrative and the combination with Kafka (the master of modern alienation) and with Vandersteen (a master of the good-humored acceptance of ordinary life) may induce Wasco’s readers to see “City” as a blend of the strange and the usual, the exceptional and the commonplace. Many things in “City” look familiar: the antennae on the roofs, the Picasso-like sculptures and paintings, the streetlights, and so on. But they always have some estranging feature: the antennae have weird forms, the sculptures are out in the street, and some streetlights look very strange indeed. The story evokes familiar schemata but disrupts them at the same time.
The mixture on the level of intertextuality is redoubled by a generic amalgam. The travel genre is evoked from the first panel of the visit multiframe. The flying saucer suggests the genre of science fiction and/or time travel. The visitors’ walk through the city reminds the reader of the typical touristic visit: one looks at all the sights, visits a museum (panel thirteen), relaxes on a bench (panel seventeen), and finally returns home (panel twenty). Again, however, this familiar scheme is disrupted: by the electric chair sitting out in the open (panel ten), the open sewer (panel fourteen), and the complete absence of any living beings apart from the two visitors. Perhaps this is a futuristic city, a city of the dead or of those living underground. Perhaps this can be seen, in line with Kukkonen, as a “renegotiation of postmodernism in contemporary culture” or more specifically as a reference to the posthuman era.125
Classical narratology developed its theory and toolkit on the basis of literary texts that were for the most part published after the beginning of the eighteenth century. Of course there are counterexamples: Vladimir Propp worked on a collection of old Russian fairy tales, and Tzvetan Todorov worked on Boccaccio’s Decameron.126 Genette’s influential conclusions about story and discourse, however, are almost entirely derived from Proust’s modernist classic, Remembrance of Things Past, creating the suspicion that the universals of classical narratology are in fact bound to a specific period in literary history. In its urge to correct what it sees as earlier mistakes, postclassical narratology therefore calls for a corpus with a much broader temporal range, incorporating texts from as many historical periods as possible in an effort to study changes that might go unnoticed if the corpus were to remain the limited set of modern books it is generally purported to be. This extension leads not only to a test of the toolkit but also allows for a deeper historical understanding of narrative as we know it today. The resulting questions are daunting, but they definitely need to be asked. Has focalization been around forever? What about the development of characterization? How has narrative evoked the reader or listener through the ages? These are only some of the issues that a diachronic narratology can tackle in its effort to debunk the grand claims of structuralism and refine the toolkit in the process.
It would be entirely wrong to suggest that narrative theory has never taken an interest in the history of narrative form. Mikhail Bakhtin, for instance, describes the development of the chronotope starting in antiquity.127 Erich Auerbach’s history of mimetic representation starts off with the elaboration of a contrast between the biblical narrative of Genesis and Homer’s Odyssey.128 In his study of characterization Erich Kahler sees an “inward turn of narrative” that finds its apotheosis in modernism.129 Dorrit Cohn’s monograph on the narrative representation of consciousness practically contains a history of free indirect discourse (or in her own terms, narrated monologue).130 Outside of narratology, period specialists have brought elements of the structuralist toolkit to their own corpora. Excellent examples for the literature of the Middle Ages include the works of Paul Zumthor and Evelyn Birge Vitz.131
Still, postclassical narratology clearly features a renewed interest in the historical dimensions of (literary) narrative. In a prominent essay,132 Monika Fludernik attributes the new motivation for diachrony to three factors: the rise of feminist narratology,133 which often insists on the importance of context for the analysis of narrative; renewed work (initiated by Michael McKeon)134 on the origin of the novel, which showed how fictional and nonfictional texts in the sixteenth and seventeenth centuries connected with each other and thus prepared the ground for the genre; and the application (inspired by Hayden White)135 of the toolkit to historiographical texts.136 Fludernik’s own Towards a “Natural” Narratology (1996) is itself an early example of the new diachronic narratology in that it retells the history of narrative fiction from Middle English prose to postmodernism through the notion of “experientiality.”137 The impact of this book on the field of narratology has also contributed to the new historicizing trend.
In her 2003 essay Fludernik homes in on scene shifts from Middle English literature to modernism, starting from the formula “Now let us leave X and Y (in location A) and turn to O and P[,] who were walking/riding/sitting in location B,” which she originally encounters in Malory and Chaucer.138 While in the Middle English period the scene shift occurs at the beginning of an episode, in the later genre of the novel it occurs at the beginning of a chapter. Since there are many more episodes in episodic narrative than there are chapters in an average novel, the result is a reduction in the number of shifts. Fludernik also finds that “the Middle English formula becomes extremely rare after the seventeenth century and is increasingly replaced by phrases with a temporal conjunction indicating simultaneity,” such as “while” and “in the meantime.”139 Furthermore, from the Renaissance onward narrators use the scene shift to put themselves briefly into the spotlight as the inventors of stories. On occasion these narrators indulge in “metaleptic somersaults.”140 An example would be “But methinks I hear the old shepherd Dorus calling me to tell you,” from Sidney’s Old Arcadia (1593).
Genette finds similar examples of metaleptic scene shifts in the works of Balzac and Proust. Fludernik, however, illustrates the huge potential of diachronic narratology by tracing the development of the Middle English formula and showing its refunctionalization for the purpose of irony and sheer narrative daring. In Orlando (1928), for instance, Virginia Woolf lets her narrator play around with the possibility of a scene shift in order to delay the announcement of the birth of the protagonist’s son.
It is obvious that diachronic narratology requires substantial historical knowledge. This constraint has at least three practical effects on research. First, there is a tendency toward period specialization; second, there are contributions that contrast older and newer literatures so as to bring out a telling difference; and third, there are collaborative projects to overcome the inevitable limitations of most scholars involved in the historicizing endeavor. We will now present examples of all these approaches in the (relatively) new field.
In a series of contributions since the 1980s, Irene de Jong has developed a history of ancient Greek narrative that has been inspired by the classical narratology of Genette and Bal. To De Jong, “ancient Greek narrative” includes “purely narrative genres (epic, novel); what could be called applied narrative genres (historiography, biography, philosophy); narratives embedded in non-narrative genres (the mythological parts of lyric, hymn, and pastoral; the prologue and messenger-speeches of drama; the narrationes of oratory); and what Genette [ . . . ] called pseudo-diegetic narratives, i.e. narratives with a suppressed narrator.”141 De Jong argues that such a broad collection is necessary for developing a plausible history of narrative form because in antiquity the same devices are found all across the corpus. Still, they may have different functions. An analepsis, for instance, may be purely informative, or it may be profoundly ideological, when the past is brought up to provide comparative material with the present, as in the work of Thucydides or Plutarch.
As De Jong sees it, Greek literature starts off with a “big bang,” since Homer “developed most of the classical narrative toolkit: the Muses, the in medias res technique, prolepsis and analepsis, embedded focalization, or the tale within the tale.”142 In Narrators and Focalizers: The Presentation of the Story in the “Iliad” (first published in 1987), De Jong proves “the serious shortcomings of the time-honoured dogma of Homeric objectivity” by considering internal (or as she calls it, “embedded”) focalization in Homer’s epic.143 For instance, in the verses “On the other side the well-greaved Greeks led Ajax, happy / on account of his victory, to illustrious Agamemnon,” the word “victory,” according to De Jong, “represents the focalization of Ajax, who rejoices about what he interprets as a victory.”144 In De Jong’s A Narratological Commentary on the “Odyssey” (2001), the close analysis of narrative devices brings to light an array of details, including the rareness of narratorial intervention and the metafictional dimension of narratorial statements about songs (which derives from the close connection between the narrator and the singers he describes, Demodocus and Phemius).145 While De Jong’s work presents an elaborate and historically sound application of classical narratology, it does not call for a revision of the universals. On the contrary, it wears its structuralist inspiration on its sleeve, and as such it has been an inspiration for many scholars in the field.
In her book on medieval narrative, Evelyn Birge Vitz is definitely critical of classical narratology. She sharply criticizes Greimas’s actantial system (presented in chapter 2 of this handbook), not least because his presentation of the relationship between subject and object is based on the vague notion of “desire.” The general ideological problem with modern narratology for the analysis of medieval texts, Vitz goes on to suggest, is that it does not manage to conceptualize “crucial medieval views and mental structures [ . . . ]. (Prime examples would be the wide implications for medieval narrative of belief in the existence and omnipotence of God, on the one hand, and of the related conviction that life largely defies human comprehension, on the other.)”146 Vitz herself may be interested in structures, but she finds the early narratological work on story grammar inadequate when it comes to the analysis of medieval narrative.
Twenty-five years later, Eva von Contzen is more enthusiastic with respect to what narratology and medieval studies can do to enrich each other. In her manifesto on behalf of a “medieval narratology,” Von Contzen acknowledges the new historicizing trend in the field overall but laments that “upon closer inspection, ‘historical’ in practice still often means post-sixteenth century, or in a huge historical leap backwards, antiquity.”147 An international medieval narratology needs to remedy this situation, but it should not do so by developing a theory of narrative informed by medieval discussions about narrative. Instead it should focus on explaining “the forms and functions of medieval practices of narration.”148
While there are encouraging signs across both narratology and medieval studies, the “alterity of medieval literature” perhaps poses the most challenging problem to the budding new subdiscipline.149 Since the period in question spans a thousand years, its integration with narratology will obviously require a great deal of differentiation, but this complication has not prevented specialists from coming up with a list of distinguishing characteristics that goes a long way in describing the thorny issues ahead. Medieval narrative questions the modern separation between author and narrator, in that the former appears in his or her own texts; it features an unusual distribution of knowledge in the sense that characters are often discussing things “about which they might well be clueless”; it offers too much or too little information compared to the norm of realistic transparency; its characters are seldom well developed; its plots include logical breaks and inconsistencies; and it creates meaning through the retelling and variation of traditional stories.150
The alterity of premodern literature holds many clues for a theory of narrative. Therefore, provided contrastive research is based on a familiarity with both epochs, it has a lot of interesting contributions to make. In a typical example Harald Haferland, a specialist on older German literature, develops the notion of retrospective motivation to differentiate forms of finality across the history of fiction.151 Events in narrative may appear to the reader as if they have been inserted to highlight a character trait. When a lover dies and his or her partner enters a period of mourning, readers might feel that the death has been written only to evoke the partner’s capacity for grief. The death of the lover is seen to be motivated in retrospect, in other words, by a later element in the narrative. Haferland argues that only modern readers would experience this as a rather cheap trick, perhaps even as a sign that the author is not capable of motivating character traits through well-organized overall composition.
To discuss the origin and form of retrospective motivation in premodern times, Haferland turns to a Middle High German rendition of the story of Pyramus and Thisbe. When Thisbe arrives early at the lovers’ secret meeting place and encounters a lion, she drops her clothes and finds a hiding place. The lion, who has just attacked a cow, rips up the clothes with his bloody jaws. When Pyramus finds the blood-stained garments, he thinks Thisbe is dead, kills the lion, and commits suicide. The fact that Thisbe needed to drop her clothes, so to speak, in order to bring about the events that follow can be seen as retrospective motivation. The narrator of this medieval version does not explain why the heroine did what she did because, according to Haferland, he is so fixed on the outcome of the story that he does not bother.
This is typical of folktales that have been retold for centuries: motivations do not matter because the teller imagines an audience that is much more interested in what happens than in explanations that go beyond the account of elementary causal connections between events. In the modern novel, on the other hand, authors have evolved to imagine a reader who is keen on precise and complete (psychological) explanations for the sake of the mimetic illusion. For proof Haferland consults Friedrich von Blanckenburg’s essay on the novel (1774).152 This is a crucial document in the history of the genre, and in it the author scolds a contemporary for not explaining why his heroine closes a door, which causes her to experience thirst all night. The contrast between modern and premodern fiction with respect to motivation proves that there are connections between narrative and narration that classical narratology would not be able to fathom because of its concentration on the later part of the canon.
In The Emergence of Mind, perhaps the most substantial joint effort in diachronic narratology to date, David Herman explains that the task of the research team was “to trace commonalities and contrasts in the presentation of consciousness over virtually the entire time span during which narrative discourse in English has been written and read.”153 The collection also includes a theoretical framework for the endeavor. In his presentation of the directions narrative research has taken to describe the experience of the reader who tries to make sense of mental activity in the fictional world, Herman distinguishes between two basic approaches. On the one hand there is the speech category approach developed by Cohn, which insists on the uniqueness of fiction when it comes to the reader’s construction of minds. Only fiction can allegedly give us an inside image of a character’s mind. On the other hand there is a body of work that questions this possibility of direct access and “gives way to a scalar or gradualist model, according to which minds of all sorts can be more or less directly encountered or experienced—depending on the circumstances.”154 This means that even famous fictional minds such as Isabel Archer’s (in The Portrait of a Lady, by Henry James) or Molly Bloom’s (in Ulysses, by James Joyce) will ultimately remain something of a mystery, but it also suggests that we can have a degree of access to even the weirdest minds, inside and outside of literature.
While the contributors to The Emergence of Mind do not always cast their findings in these terms, Herman’s initial presentation suggests that the discursive evocation of mind has a less dramatic history than the proponents of an “inward turn” in modern narrative might like to make it seem. It is certainly not the case that there is no early sophistication. In the chapter on Old and Middle English narrative, for instance, Leslie Lockett argues that the texts in her corpus are shaped by conceptions of mind that are nondualist and intrinsically related to the body. Intense mental activity is described as the mind heating up and growing in size in the chest region or, even more precisely, in the heart. Since there was no constant exposure to Neoplatonist-Christian psychology or to the findings of Galenic medicine (both practices that “helped to relegate the corporeal mind [ . . . ] to the status of metaphor in Modern English”), these descriptions “were intended or interpreted by their target audiences as literal articulations of the mind-body relationship.”155 In other words, fictional texts like Beowulf mirrored the mind as it was known at the time and thus reinforced its dominant conception.
Elizabeth Hart in her contribution to The Emergence of Mind concentrates on three late sixteenth-century romances, by Sir Philip Sidney, Robert Greene, and Edmund Spenser, to document the move away from the premodern link between heart and mind to the modern nexus between head and mind. Hart does argue for an important change in the evocation of consciousness from premodern to modern, and she connects this development with the “massive shift in people’s cognitive abilities toward reading in particular” that followed the invention of print technology.156 The growing readership meant that people moved away from “thinking about things to thinking about representations of things, that is, thinking about thought.”157
In the romances under consideration, Hart finds two early indications of the new representations of mind in direct reference, “the moment when the narrator zooms in on the head or mind of a represented human and thus objectifies it as a container of interiority,” and (unsophisticated) thought simulation.158 Here is an example of the latter from Sidney’s Arcadia: “Then turned he his thoughts to all forms of guesses that might light upon the purpose and course of Pyrocles, for he was not sure by his words that it was love as he was doubtful where the love was.”159 Her insistence on a major shift does not mean that Hart pleads for the uniqueness of narrative fiction when it comes to the evocation of mind. On the contrary, the signs of novelty she detects in literature are part of an interaction between the new interest in mind and the texts that help to feed it.
David Herman’s own contribution to The Emergence of Mind undermines the classical image of modernism as a period in literary history in which “the accent falls less on fictional worlds than on fictional-worlds-as-experienced.”160 This shift may be borne out by the increasing use of internal focalization, the development of interior monologue, and a focus on mental incapacity (as with the character Benjy in Faulkner’s The Sound and the Fury), but Herman submits that the advances of modernist fiction can better be described with reference to “postcognitivist” theories of the mind. From this perspective the famous novels of modernism, such as Henry James’s The Ambassadors or Virginia Woolf’s The Waves, would become attempts “to highlight how minds at once shape and are shaped by larger experiential environments, via the particular affordances or opportunities for action that those environments provide.”161 Zooming in on a passage from James Joyce’s A Portrait of the Artist as a Young Man in which the protagonist Stephen Dedalus goes to confession, Herman shows that the text is first and foremost based on Stephen’s routine knowledge of this activity, as it includes opening the door to the confessional, kneeling, and saying the required formulas. In its turn, this construction of the environment leads to a set of perceptions, inferences, and other reactions that determine Stephen’s “navigation” of the surroundings.162 Herman finds similar “action loops” in a short passage from Woolf’s Mrs. Dalloway.163 That text combines two perspectives on the arrival of the British prime minister at a party thrown by the title character. Both responses derive from what the situation affords—in the case of the female character, the possibility of telling her friend, and in the case of the male character, the possibility of developing thoughts that might prove Mrs. Dalloway wrong in having chosen a rival over him.
Insights derived from diachronic narratology could help to analyze narrative texts, even if the focus is limited to a single text. Herman’s reading of Mrs. Dalloway sheds new light on the interaction of Woolf’s characters with the environment, and it illuminates new thematic accents in the novel. Along the same lines, one can read the evocation of consciousness in “Pegasian” through a historical lens and find new meaning in the process. The presentation of the two competing perspectives on the riding breeches could have taken the form of direct quotation in a simple dialogue between the two protagonists—the riding master and the “little girl.” There is of course quite a bit of direct quotation (e.g., “Do they make you go faster? / No, not faster”), albeit without quotation marks, but it is caught in a frame of indirectness that easily allows for access to the minds in action. When the girl has asked her question about breeches in the form of fat, the riding master retreats into personal thoughts, which leads to a brief interior monologue. Similarly, in the final paragraph the narrator reports that the girl understands, and the reader immediately hears the explanation she uses to convince herself.
Historically speaking, “Pegasian” author Charlotte Mutsaers displays a great mastery of the various techniques for evoking the mind at work that have been developed from the end of the eighteenth century onward. In fact her story amounts to a small catalog of these devices, from direct quotation (spoken thoughts or interior monologue) over free indirect speech to indirect speech. Perhaps the narrator’s opinion ultimately coincides with that of the girl, but the ease with which the narrator moves from one device to another can be brought into consideration as proof of a flexibility that the protagonists at first seem to lack and that the riding master never attains. The narrator is able to move back and forth between the various ways to represent consciousness, but the two main characters are stuck in their own convictions. If at the end of the story the narrator’s opinion is seen to coincide with that of the girl, the latter could also be said to have reached the mental litheness the narrator has showcased all through the text. “Pegasian” could not have been written like this until the modernist completion of a long process in the evocation of mind that includes, among many others, the works of Jane Austen and Gustave Flaubert. Through its historical dimension, the story’s formal variation can be validated as a counterpoint to its theme of obstinacy.
From a diachronic perspective “The Map” is a typical example of the first-person narrative perfected at the latest by Charles Dickens in David Copperfield. In examples of this kind of text individuals look back on their past with a certain amount of insight, mostly relating the events of their lives in chronological order and often judging them in the process. Gerrit Krol does not quite break new ground with his story in terms of the formula, but the classic structure of “The Map” can still be brought to bear on its contents. The boy who was so keen on acquiring the map and employing it “to be everywhere” has become an adult for whom the object has become “useless.” The narrator clearly states that his dream of being everywhere “fade[d] away” the more he traveled. He completely lost interest in and even discarded the map. This anticlimactic action suggests that the man has lost the excitement of youth, but Krol’s use of an old formula might be read as a complication. His choice of an archaic formula implies a degree of (unspoken) nostalgia, whereas the abandonment of the map denotes the exact opposite. Unless the reader interprets the final sentence of the story as a telling variation on the elements of peace and maturity as they characterize the situation of the typical older I-narrator, this contrast remains unresolved.
In classical narratology the world in which the narrated actions take place is studied in terms of space, time, and agents. These are investigated as abstract categories, elements in a self-sufficient system and structure. The link and comparison with the world outside the narrative is not deemed important, if it is mentioned at all. Postclassical theories broaden the fictional world by linking it with social reality (through the study of the fictional world as a possible world) and with the world of the reader (by considering the fictional world as the reader’s construction). We will first look at possible world theory and then at the storyworld approach proposed by David Herman.
The “possible worlds” concept, which was introduced in narratology as early as the 1970s, derives from modal logic. This discipline investigates the possibility, impossibility, contingency, or necessity of propositions. Ruth Ronen,164 Thomas Pavel,165 Lubomir Doležel,166 and Marie-Laure Ryan167 in particular have pointed out the usefulness of this concept for literary theory. In order to characterize the modal structure of a literary text in a more concrete way, we will take a close look at Doležel’s classic, Heterocosmica. In a further exploration of possible world theory, we will discuss Uri Margolin’s contribution to the study of character and David Herman’s work on hypothetical focalization.
Possible world theory studies the world of fiction along the same lines as any other world, including social reality. “The basis of PW [possible world] theory,” writes Marie-Laure Ryan, “is the set-theoretical idea that reality—the sum total of the imaginable—is a universe composed of a plurality of distinct elements, or worlds, and that it is hierarchically structured by the opposition of one well-designated element, which functions as the center of the system, to all the other members of the set. The central element is commonly interpreted as ‘the actual world’ and the satellites as merely possible worlds. For a world to be possible it must be linked to the center by a so-called accessibility relation. Impossible worlds cluster at the periphery of the system.”168
This triadic model distinguishes between the center of the world (the actual world, the existing state of affairs), the possible worlds circling this center as satellites, and the impossible worlds at the outer limit. The difference between satellites and the periphery lies in the so-called accessibility relation that they have with the center. Possible worlds have access to the existing worlds: they could at one point become real. How this is determined differs from one theoretician to another, but often the laws of logic and time are used as criteria for deciding whether the literary world can gain access to the real world or not. A world in which the logical law of the excluded middle is not respected (for example, when a character is at the same time dead and alive) will be called peripheral or impossible. In this view a story in which an old man suddenly becomes a child would create an impossible world as well. Looking at our three stories, it is obvious that the ones by Mutsaers and Krol are closely linked to our own actual world, whereas Wasco’s world is far removed from it, though in some future it may become an actual world. On its own, each story world has a center, a set of satellites, and a periphery. The centers of the worlds of Mutsaers and Krol are much closer to ours than is the center of Wasco’s futuristic world.
By studying narrative and social worlds in terms of the same triadic model, possible world theory sheds light on the autonomy and the relatedness of fictional worlds. The fact that fiction as a possible world constitutes an autonomous, closed system means that it can be compared without any problem to another possible world (such as dreams) and that it can even be described by referring to that other ontological system—the existing world to which it relates. Like Umberto Eco, many literary theorists tend to interpret the existing world as a cultural construction.169 Fictionality is then conceived as the result of the interplay between the system constructed by a literary text and the system available to authors and readers in the form of knowledge of the existing world. The world created by a novel is called fictional because it is seen as an alternative to the existing state of affairs.
For the theory of narratology the most general advantage is the presentation of a philosophically sound framework that considers the narrative text as a system that has its own laws and is at the same time related to the context. In a classic detective novel the resurrection of a character who has been murdered is impossible. In this case the definition of what is possible corresponds rather well to what is considered to be possible in the human experience of reality. In a nonclassic detective novel, like Atte Jongstra’s Het huis M. (The house M), such a resurrection turns out to be perfectly possible.170 The murder in question is committed over and over again as well. What is possible in this novel clearly deviates from what is possible in everyday reality.
To a considerable degree, what is possible in narrative depends on the reader’s assessment. At the beginning of Louis Ferron’s novel De keisnijder van Fichtenwald (The Fichtenwald stonecutter), the character Friedolien sees the environment as peaceful and comforting, while the reader immediately notices that it is a concentration camp. His or her knowledge of these camps will gradually reveal Friedolien’s problematic observation as a lie.171 David Herman gives the example of a statement concerning Brussels. If a character in a text says that Brussels has many interesting museums, this does not automatically mean the character asserts that the capital of Belgium has interesting museums. Readers who know that Brussels is the capital will, however, probably interpret the remark in this way, and thus they will create a possible world that may very well deviate from the character’s possible world. Every reader has a different kind of knowledge and therefore constructs his or her own possible world.172
The concept of “possible worlds” can be related to “virtual reality,” an idea that often appears in the study of hypertext. According to Ryan, virtuality has three dimensions. First, the term can be considered a synonym of “illusory.” Virtual reality is a feigned reality that gives us the illusion it is real. Second, the concept refers to computer technology. Virtual reality is the world evoked by technology, such as the World Wide Web. Third, virtuality may be synonymous with possibility or potentiality. In that reading, virtual reality becomes a potential or possible world, and in this way cybernarratology can be related to modal logic and the narratology of possible worlds to which it has given rise.173 Just as cybernarratology sees the textual world as a palimpsest that comprises different layers of virtual realities, possible world theory envisages the world as a composite of potential and existing realities.
From Plato to Bertrand Russell, conventional theories of truth determined the truth value of a proposition by looking at the correspondence with a situation in the world. Therefore, propositions in fictional texts did not have any truth value whatsoever. Nevertheless, literary texts often do refer to reality. In order to characterize this type of reference, the philosopher John Searle described fictional speech acts as making it seem as if they refer to reality. They do not have to meet all the requirements of a normal referential speech act. In this way Searle lifted the usual rules of truth for literary texts, and in a certain sense he thereby rehabilitated this type of text.174
In pragmatic theory, truth is no longer considered the result of the connection between language and world. Moreover, truth is no longer seen as a matter of everything or nothing. To the reader or hearer, statements can be acceptable and plausible, and in that sense they can be truthful. According to this theory, a proposition can be true in one way or another, even if that which the proposition refers to does not exist. For a certain utterance to be true, the state of affairs that is referred to does not have to exist in reality. Even if one does not know whether the state of affairs is real, the utterance can be considered true. This depends on the extent to which the utterance is judged to function well in its context. If utterances “work,” they are accepted as true. Utterances in a novel that contribute to the plausibility or impressiveness of the book could be considered true. They work perfectly within their context, and that is enough.
This conception of truth corresponds nicely with the antimimetic theory of fictionality, which argues that fiction creates its own discursive universe in which propositions can be true or false. Internal criteria are sufficient to reach a decision. However, this should not be confused with the traditional view on fictionality as an inherent characteristic of fictional texts. Those texts are traditionally seen as having intrinsic, intratextual signposts of fictionality, such as free indirect discourse and the distinction between author and narrator. Ronen explicitly distances her approach from these so-called “textual-taxonomic models of fictionality.”175 Whereas the textual models focus on semantics and syntactics, possible world theory takes a broader, pragmatic view in which fictionality is a particular form of creating a world, implying a specific relation between sender and receiver and between textual and real world. “Fictionality,” Ronen writes, “has to do with the relation between a speech situation and its context, and with the degree and kind of commitment of the speaker to the content of his utterance. [ . . . ] An alternative perspective on the problems involved in defining fictionality would stress that the fictional by definition does not refer to an inner structure but to a type of relation: a relation maintained between what is contained within the literary text and what lies beyond its boundaries.”176
The world evoked by the literary text may deviate from reality, but this does not mean that it is an untrue world. Moreover, this perspective on fictionality does not imply that literature is irrelevant for extratextual reality. Fictional texts always function in relation to nonfictional texts and can therefore tell us a lot about these so-called realistic texts, as well as about so-called reality. For example, the relationship between historical novels and historiography can teach us much about the ways in which we deal with the past (the so-called extratextual reality).
Apart from this general and context-oriented relevance, possible world theory also enables narratologists to envisage the narrative world within the literary text as a collection of possible worlds. The construction and evolution of the story are seen as the interaction between those worlds. Every possible world within a single text is in principle defined by means of a specific modality such as probability, possibility, or necessity. In “The Map” the boy sees the desire to map everything and to have been everywhere as among the possibilities. For the adult narrator this desire forms part of the impossible world, and the failure of this desire is a necessity. The interaction between the different possible worlds appears to be essential for the development of the character and the story.
The interaction and the relationship between the worlds differ from story to story. In a naturalist text, that which is possible mostly has to surrender to that which is necessary: dreams clash with unchangeable reality. In Willem Brakman’s novels this hierarchy is often reversed since the main characters prefer impossible dreams to achievable realities. Coercive reality is a horror, possibility leads to disaster, and only the impossible is interesting. The main character in De sloop der dingen (The wreckage of things) aims for the impossible suspension of time. The demolition of his village entails a modal statement: “I consider it a disaster that this Duindorp will be demolished, and therefore it is possible.”177 Reality is the result of this demolition—that is, degeneration and death—“which would make everything lethally real.”178 The different modalities alternate infinitely, without there ever being a final victory or resolution, so the interaction of fictional worlds in Brakman’s novel does not lead to the completion found in the work by Krol or works by the naturalists. The reader is offered a story without a straightforward plot or clear ending. The endless alternation of possible worlds lies at the basis of this seemingly directionless story that nonetheless aims to put off the necessary ending and the real demolition as long as possible.
In order to study the different kinds of modality in a novel’s narrative or fictional world, Doležel has developed a four-dimensional system.179 He characterizes every dimension by means of three terms. First, the narrative world can be described from the perspective of alethic modality—from the Greek word aletheia, meaning “truth.” Alethic modality refers to everything that is necessary, possible, or impossible according to the laws of nature and logic. Necessity, possibility, and impossibility constitute the decisive criteria for alethic modality.
The clearest examples of this are causality and spatiotemporal specificity. A fictional world in which people can fly violates the laws of nature and therefore constitutes a supernatural world. Physically speaking, this world is impossible. However, it may very well be logically coherent and, in that sense, logically possible. The world becomes logically impossible only if logical laws are violated in this supernatural world as well. A fairy tale in which people can fly does not have to be a logically impossible world. Moreover, alethically speaking, many intermediate forms are possible. Between the natural and the supernatural world there are intermediate worlds—such as dreams and hallucinations—that may be explained in a perfectly logical and natural manner, as, for instance, when hallucinations are triggered by the use of drugs. That which is impossible in one world and for one character may be possible in another world for another character. This situation plays an important role in the definition of the hero, who in most cases is capable of doing things other characters cannot.
A second form of modality in the narrative world has to do with norms. In a fictional world certain things are prohibited, others are obligatory, and still others are permitted. Prohibition, obligation, and permission are the three building blocks of deontic modality. According to Doležel, the deontic marking of actions is the richest source of narrativity.180 An action—for instance, a trip—can appear perfectly neutral in and by itself, but deontically speaking it may turn out to be a violation of a prohibition that may trigger an entire system of counteractions. Typical narrative patterns such as the test, the initiation, and the fall can be analyzed from the perspective of this modality.
Of course these norms may change in the course of the story, and characters may play an important role in this change. Their importance and status often even depend on their contribution to the transformation of dominant norms. Heroes may demonstrate their power by determining what is permitted and what is obligatory. The world of norms is in constant development because norms are the stakes of a constant struggle. The struggle between personal and general norms forms the basis of stereotypical narrative patterns such as forbidden passion or liberation from a stifling environment.
The third modality for describing the fictional world is axiological modality; it deals with moral judgment. In this case as well there are three possibilities: good, bad, or indifferent. There is a constant interaction between subjects (the characters) and their environment. That which characters think of as good can be bad according to their environment. Just like deontic modality, the axiological dimension forms an important source of narrative actions. In most cases axiological modality leads to actions via the detour of desire or repulsion. Traditional heroes will desire what is good, and this desire will stir them to action. Nihilistic characters appear driven by indifference: they find the values of their environment neither good nor bad; such characters do not care for them.
The fourth and final modal building block of the fictional world is the epistemic modality, which consists of three possibilities: knowledge, ignorance, and belief. The last refers to presuppositions of characters that are not based on the real state of affairs in the story. Knowledge is mostly distributed unevenly in the characters, and this forms an important source of narrative actions. A detective is the best example of a character starting from ignorance and carrying on to arrive at knowledge. A schemer can be said to exist thanks to the ignorance of victims of the schemes. Misunderstandings and false presuppositions are at the basis of typical narrative patterns, such as in a comedy of errors. More generally, the interaction between knowledge, ignorance, and belief is central in narrative patterns, as in a quest, deception, and disappointment.
The action of a story can easily be studied as the interaction between the four modalities discussed above. What is possible may be, for instance, forbidden and bad. As long as this possibility does not penetrate a character’s consciousness and knowledge, there is no problem. But when the character learns of this possibility and starts seeing it as valuable or good, all kinds of narrative possibilities crop up. A character facing an explicitly formulated prohibition may become conscious of something related to that prohibition, which may paradoxically result in a violation of the latter. An extreme example of this can be found in the following promise: “You can have the treasure that is buried under this tree, but you must not think of the big bad wolf while digging for it.” Without the prohibition, the thought of the wolf would probably never have entered the hearer’s consciousness, and the prohibition would never have been violated.
The interaction between the four modalities can explain many aspects of narrative development and plot. This way, the structuralist conception of plot can be improved. Classical narratology usually determines narrative structure starting from knowledge of the ending. This structure therefore tends to be described by referring to events that have actually happened in the fictional world. Nevertheless, things that have not really happened (for instance, dreams and plans) are often essential for narrative structure. If the narrative text is considered an interaction of possible worlds, so-called nonevents come to the fore. In this kind of interpretation, one takes into account modal aspects that do not occur as real events in the story but are very important for the meaning of the story, as we have discussed in the second chapter of this book.181 The continuous impossibility of a certain situation cannot be ignored, because it can be of vital importance for the things that are possible or even real during that period. Madame Bovary is largely driven by the main character’s unfulfilled desires, so the conflict between the possible world and the actual situation she finds herself in has to be part of the narratological interpretation of Flaubert’s narrative structure.
Modality is also essential in the possible world approach of fictional characters. Uri Margolin’s renowned analysis of character finds its “theoretical foundation” in the combination of modal logic with possible world theory: “Modal logic is basically the study of what is to be considered possible or necessary in some world[,] while possible-world semantics is the study of alternative worlds, their governing laws, and the kinds of individuals inhabiting them.”182 From this perspective a character is studied as an individual (IND) in a nonactual world. Margolin puts it this way: “In this view, a narrative is a verbal representation of a succession of hypothetical states of affairs, mediated by actions or events. The IND is a member of some domains of this possible world, and in them, it can be uniquely identified, located in a space/time region, and endowed with a variety of physical and mental attributes and relations.”183
As such, any character can be located in time and space and can be characterized in terms of three dimensions: physical, mental, and “behavioral (action-related) and communicative.”184 This makes the character look like a human being, but there are basic differences too. For one, characters exist only in the words devoted to them: “In fact, they are these complexes of descriptions, not having any worldly existence.”185 As a result, they are never as complete as a “real” person. They are “ontologicallly ‘thin’ and not maximal, having only a limited number of properties and relations. Unlike actual INDs, they are schematic, radically incomplete, and only partially determinate.”186 Real persons may function as characters in a fictional world, for example, in a historical novel about Napoleon or Dante, but even then they are reduced to textual and incomplete existence. Umberto Eco talks about “transworld identities,” Philippe Hamon about “referential characters”; to Margolin, they constitute the “actuality variant,” as opposed to the category of completely made up characters, “the supernumerary.”187
Characters can be recognized only if they have a core that remains more or less unaltered. At the same time, however, they change, slightly or drastically, with every reappearance in the storyworld. Margolin studies the static dimension of character in terms of four conditions. Three of these are necessary: “existence, individuality, distinctness or singularity”; the final one is optional: “paradigmatic or simultaneous unity of traits.”188 Each character meets these conditions in a different way and to a different degree, thus allowing for the different types of characters proposed in traditional narratology (e.g., round versus flat).
“Existence” means that the character functions as an entity in one or more of the possible worlds constructed by the narrative: “The existence condition is satisfied for a given IND if and only if the text establishes uniquely, stably, and unequivocally its membership in the narrated domain or any of its subdomains.”189 There are four domains in which the characters may feature: the textual actual world, which is established by the highest-level narrator and which constitutes “reality” in the narrative; the “counterfactual, hypothetical situations,” which run counter to that “real” world; the “subjective (wish and belief) domains of INDs in the narrated domain,” including a character’s interpretation of the “real” world; and, finally, “unrelated subdomains representing stories these INDs read or hear, movies they watch, and so forth.”190 For instance, in “The Map” the boy’s reading of the textual actual world is channeled by the map, which is at the same time an almost scientific, objective rendering of the real world and a sign of the boy’s subjective wish to cover that world by biking. His evolution entails a shift in the importance of these two aspects: his dreams fade away, and the world becomes smaller, less relevant.
The second condition that characters must meet is “individuality [ . . . ], the ascription of properties to an individual picked out by a referring expression.”191 These properties pertain to the three dimensions mentioned earlier: physical, mental, and behavioral (action-related) or communicative. They tell us what the individual is like (whereas “existence” simply told us that it existed), and they do so in direct and indirect ways (which rehearses Shlomith Rimmon-Kenan’s proposal about characterization, discussed in chapter 2) and by descriptions that may be unique to the IND (e.g., name) or generic to the IND’s type (e.g., mortality).192
One step further in the characterization leads us to the third condition: uniqueness or singularity. This means that “each IND must at each point be different from all its world mates in one or more properties.”193 Central characters tend to have a vast amount of what might be called distinctive features, whereas background figures tend to have fewer. Singularity is not just about distinction, but also about relatedness. Characters can be related and compared to others, again in terms of the three dimensions (mental, physical, action/communication). In “Pegasian” the female figure is clearly distinguished from her co-riders: she is an inquisitive “little girl” as opposed to the “ladies” who never ask questions. In “City” the central character could be the only humanoid form left in a completely artificial and inanimate environment.
A final, and optional, condition concerns the possibility of turning the set of characteristics provided by the first three conditions into a unified whole—a class or type. Margolin speaks of “a global configuration or a unified properties and relations complex.”194 This configuration takes places on the basis of three sources: “intratextual, explicitly formulated models in the discourse of the narrator; intertextual models of named INDs (Don Juan), or recurrent generic models (the suffering artist, the demonic personality, the seer); and extratextual models of an age or society stemming from their encyclopedia (in Eco’s sense) or explicitly formulated doctrines.”195 In Ivan Turgenev’s story “Asya,” Gagin, the half brother of the title figure, is described as an aristocratic would-be artist, both by himself and by the I-narrator of the story. This characterization may be labeled intratextual, but it also activates the reader’s encyclopedia and therefore comes down to extratextual characterization as well. The same goes for the following description of Gagin: “he literally exuded all the softness and near-effeminacy of a true member of the Russian nobility.”196 Again, possible world theory studies elements of the narrative world in their relation to the “real” world of the author and/or the reader.
In addition to this static, fourfold aspect of the character, there is the dynamic dimension. Change can be studied on the level of the character and of the world in which it figures. The two are intertwined, but they can be distinguished to keep the analysis simple. As to characters, their changes may be global and coherent (following a clear logic and leading to a clear outcome, as in the case of someone becoming ever more corrupt and evil) or local (involving different forms of changes and not following a predictable logic). Furthermore, changes may be small or big, as well as gradual or sudden. Using these parameters, one might, again, distinguish and classify different types of characters in a way that would be far more refined than the traditional distinctions between static (flat) and dynamic (round) characters. In changing, characters affect and are affected by other characters. This may take extreme forms, such as in two characters exchanging their personalities or in one character being cloned.
Second, character dynamics can be described as an evolution in possible worlds. If a character becomes paranoid, this development may be analyzed as a transformation of the subjective or even counterfactual world into the textual actual world: what is “merely” the world of wishes and fears is described as the real world. That would count as an example of an intratextual change. Worlds can also change on the intertextual level, as in the TV series Elementary, in which Sherlock Holmes is transformed into a twenty-first-century sleuth and his companion becomes a female doctor, Joan Watson. Both the world and its characters are transformed. Again, this may go hand in hand with extratextual metamorphoses, for example, in the transformation of stereotypical social figures or of real, historical persons. The comparative study of the intratextual character and its inter- or extratextual counterparts may draw attention to all kinds of alterations, such as subtraction, addition, substitution, and rearrangement.
Margolin’s model allows for a subtle, in-depth analysis of characters and characterization. It elucidates static and dynamic aspects, and it provides a clear model for the intratextual, intertextual, and contextual study of character, paying equal attention to textual description and readerly processing. Moreover, it can easily be combined with the classical investigation of direct and indirect textual characterization. As such, it is an excellent example of postclassical narratology enhancing classical studies.
In the construction of a possible world, focalization plays an important role. A certain type of focalization implies a certain degree of (un)certainty and (im)possibility. If, in De keisnijder van Fichtenwald, by Ferron, the observation of the camp as a resort came from a reliable narrator instead of from a mendacious character such as Friedolien, readers would consider the observation to be more trustworthy. Perhaps they would conclude that in this novel things beyond Friedolien’s world are possible. When a reader accepts a novel’s nonrealistic world as real or possible, Marie-Laure Ryan talks about “recentering.” Normally the actual world is the world of which I am the center; I cannot move outside it, and therefore I cannot consider it just a possibility.197 If readers are convinced—or carried away—by the text they are reading, they surrender their outsider position and place themselves in the center of the fictional world, which is thus moved from (im)possibility to reality. The text’s focalization is one of the elements directing this change in the reader’s viewpoint.198
But such a form of recentering does not always lead to a long-term or permanent acceptance of the textual world as real. In many cases one recentering follows another, just as one focalization follows another. An unequivocal conclusion about the reality of the represented world is not always within reach. Narrative texts show different degrees of certainty with regard to what is being narrated, and sometimes it is impossible to draw a clear boundary between what the text sees as actual and what is conceived of as a possible world. This is certainly the case in what David Herman calls hypothetical focalization. The term refers to an impossible center of observation. Inanimate things such as stones and books become centers of experience in Edgar Allan Poe’s “The Fall of the House of Usher.” In Een fabelachtig uitzicht, Gijs IJlander lets a stuffed squirrel observe the action and tell the story. What is made up and told by the dead animal becomes reality in the life of Zaalman, the main character and the taxidermist who prepared the animal. Hypothetical observation is not always tied to a nonactual observer but can be built in via conditional constructions as well, as in Salman Rushdie’s Haroun and the Sea of Stories: “Miss Oneeta was standing on her upstairs balcony, shaking like a jelly; and if it hadn’t been raining, Haroun might have noticed that she was crying.”199
David Herman combines possible worlds with focalization by placing the structuralist focalization types on a continuum. At one end there is total certainty about what is being communicated, and at the other end there is total doubt. External focalization implies a distance that can be interpreted by the reader as a signal of certainty. He or she may be wrong, as for instance when an uncertain or unreliable narrator uses a so-called omniscient perceiver. Internal focalization moves in the direction of doubt. In the case of a fixed internal focalizer, attitudes and opinions are in most cases rooted in a single possible world, and this may induce a feeling of uncertainty in the reader, especially when the character’s observations do not correspond to what the reader thinks is normal. Nevertheless, a fixed center of observation generally raises less doubt than variable and multiple internal focalization. These focalization types typically produce a pair or a whole series of possible worlds, often without any clear coherence or systematization.
David Herman places hypothetical focalization closest to the pole of doubt. With this type the existence of the observer is uncertain, which inevitably raises doubts concerning his or her observations. However, in our opinion a reader may also recognize this type as a very conventional way for a narrator to introduce new elements without raising doubts as to their value. Eighteenth-century novels are crammed with omniscient narrators who from time to time try to make their omniscience more credible by introducing reservations about their own statements. The repeated use of words like “perhaps” and “probably” gives the reader an impression of reliability rather than doubt. Hypothetical focalization can create this impression as well. Fragments with observations by an alter ego, a double, or a phantom may confirm rather than undermine the main character’s experiences.
“Pegasian” sketches a fictional world in which human beings can fly on a horse. This feature is a possibility that is not really put into question in this fictional world. The question posed by the story is not, Can a human being go up in the air on a horse? The question is rather, Do the riding breeches contribute to this form of transcendence? This problem is the focal point in the struggle between the girl and the riding master. The conflict does not in the first place concern the truth or falsity of a world in which people can fly; it concerns the way in which this world can be reached. Nowhere does the riding master deny that one can go up in the air on a horse. This instructor merely denies that this is possible without the necessary discipline.
The riding breeches play a crucial role in this self-training. The discussion on the importance of the trousers is emphasized by a variable focalization that remains close to the spoken dialogue. The riding master is convinced of the utility of the riding breeches, while at least in the first part of the text the girl thinks they are superfluous. The difference between the two figures can be described using Doležel’s modalities. With respect to the axiological dimension, the riding master thinks the riding breeches are good, but the girl does not. At the deontic level, the riding master maintains that the breeches are obligatory, but this does not hold true for the girl. Alethically speaking, the breeches are necessary according to the riding master, but in the girl’s opinion they are not. In the end the battle moves to the third pole of the dimensions. Axiologically, the pole of indifference wins out: it does not matter whether one wears those breeches or not. At the deontic level the pole of “permission” wins out: the riding breeches may be worn, but “a simple straightforward denim pair” works equally well. Alethically speaking, the riding breeches are part of what is possible; they do not belong to that which is necessary. Looking at the resolution from the perspective of the shifts at these three levels, the story eventually chooses the greatest openness possible: the “winning” poles are “possible,” “allowed,” and “indifferent.” In this way the struggle between the two characters demonstrates that many things are possible in the fictional world of this story—probably more than in the world the reader considers to be real.
There is yet a fourth modality: the epistemic one. Looking at it from this dimension, one sees that the story starts from the struggle between a character pretending to know how things ought to be done—the riding master—and a character who apparently knows nothing and constantly asks questions—the girl. At the end the girl does know how things work (“Finally she understands”), but this knowledge turns out to be a form of ignorance. She does not know at all whether ideas or sensations are at stake, but she does know that does not matter. Her knowledge does not amount to a servile acceptance of the insights offered by the riding master, who seems to know everything and who would be able to choose between idea and sensation (“it’s rather the sensation that matters”). The girl’s knowledge consists of acceptance of a certain kind of ignorance.
All things considered, the elusiveness of Mutsaers’s story could be explained from this perspective as a combination of modalities that are usually kept separate: knowledgeable and ignorant, good and indifferent, obligatory and permitted, necessary and possible. This combination makes a variety of things possible in the story, and it also explains why many elements are left up in the air—literally as well as figuratively. The confrontation of different opinions about what is possible and what is not leads to a combination rather than a selection or a choice.
The “struggle” pertaining to this confrontation is not a noncommittal or abstract display of different possible worlds. The possibilities are not totally free or God-given at all. They come about and they are imposed in a context of authority. The riding master has an authoritative position, perhaps thanks to age (the instructor is, in any case, older than the girl), perhaps also because of experience and knowledge. The possible worlds of the two characters are strongly influenced by the riding master’s position.
In the first part of the text the two characters differ not only with respect to their attitude toward the riding breeches but also in their views on authority. The riding master considers the authority conveyed by employment as an instructor to be self-evident. This shows through in the quiet self-confidence with which the master tries to persuade the pupil to wear the breeches: “The riding master would appreciate it if she’d remember.” The pupil, on the contrary, is not impressed and has doubts concerning the riding master’s recommendations, partly perhaps out of an adolescent dissatisfaction with the power of people who are older. When the riding master refuses to budge from an opinion already expressed and then behaves more condescendingly, the girl becomes more reckless. It does not look as if their ideas about what is possible and what is impossible will ever coincide. The end is not the final outcome of the struggle between the riding master’s authoritarian attitude and the antiauthoritarian manner of the girl. It is very well possible that both attitudes allow one to take off. At the level of authority all possibilities remain open as well, and the story refuses to promote one possibility to the status of the “real” state of affairs.
Simplifying a little, we could say that classical narratology could be said to limit itself in most cases to that real state of affairs. It nearly exclusively pays attention to the so-called factual building blocks of the narrative world, such as events, characters, and setting. The nonfactual can never be approached in a concrete way. Possible worlds narratology provides a theoretical framework in which the nonfactual can be analyzed in a detailed manner, that is, as the interaction between various modalities.
In “The Map” the title object constitutes a world in itself. It opens up an almost infinite possibility that is opposed to the actual world of the village, which is dominated by constraints and prohibitions. There are mechanical and ritual regulations for the purchase of a book, and there is a prohibition against purchasing the map on Sunday. And on Monday the boy also could not buy the map: “I did not have enough money, so that I had to wait until Saturday.” Obligation and nonpermission are the crucial modalities in the world of the village. Deontic modality pushes everything else to the background.
The map intensifies the experience of the first-person narrator to such an extent that he adapts his experience of reality to it. He wants to make his world coincide with the map—first with that of the village, later with “a blank map of the Netherlands.” Literally and figuratively, he transgresses the village borders. Deontic modality is not so important here, since the boy does not seem to be interested in the violation of a prohibition. He is driven rather by an epistemic desire, a pursuit of knowledge. He wants to get to know the world and map it. Thus the map becomes “[a] whole table full of new things.” This world is literally an outside world, a domain outside the realm of the village. The boy aims to integrate the outside world into a system, a map. That is what he thinks is good. At the axiological level one could say that “good” is linked to “knowledge,” and this knowledge would then be a question of mapping.
The first-person narrator does not always perceive this overlap of two possible worlds—the map and the areas he wants to visit—in the same way. When he is telling his story, the map has lost much of its significance, and his world is no longer oriented to coincide with this map. The narrator has imported the outside world almost entirely into a system, and this is precisely the reason why that world has lost so much of its attraction. The interiorization of the outside world is a desire the young experiencing “I” pursued only temporarily. The map led to an expansion of his horizon and his knowledge. The young boy’s dream starts to fade, however, as soon as he travels to farther destinations more easily and the dreamed-of expansion starts to belong to his actual world. Knowledge is no longer identified with “good” but rather with “indifferent”: “It had become meaningless.”
It is important to keep track of the narrative situation in this interpretation. When he tells his story, the first-person narrator has already been through the whole process of the expansion of his horizon, and this undoubtedly colors his representation of the world before that expansion. His focus on the shading of the Christian shops’ windows, for example, is directly linked to liberation, which is one of his central themes. The quotation of the opinion on the Paalman couple suggests a normality propelled by gossip. Their routine interaction with a customer indicates the absence of excitement. The actual world of the young boy is clearly distorted by retrospection, that is, by the actual world of the adult narrator. To formulate matters in strictly narratological terms: his focalization as a young member of the village community is determined by his focalization as an adult. The possible worlds of the map and the desire to expand his domain can be seen as the result of the projection to which the adult first-person narrator surrenders in his memories.
As David Herman makes clear, his notion of storyworld is inspired by efforts on the part of possible worlds theorists such as Doležel, Pavel, and Ryan “to overturn the structuralist moratorium on referential issues.”200 Herman’s interest in the readerly assembly of fictional worlds (which we will further discuss in the following section on communicative approaches to narrative) leads him to describe the topic of narrative analysis as “the process by which interpreters reconstruct the storyworlds encoded in narratives.”201 A contraction of “story” and “world,” storyworld has turned out to be a most powerful coinage in postclassical narratology, to the point that an academic journal on narrative has been named after it.202 As a concept, it is more specific than the phrase “fictional world,” since it stresses the cognitive dimension in the world’s construction by the reader.
Storyworlds, according to Herman, “are mental models of who did what to and with whom, when, where, why, and in what fashion in the world to which the recipients relocate [ . . . ] as they work to comprehend a narrative.”203 The creation of such a mental model is absolutely fundamental; without it, there is just no way to make sense of a story. Herman slightly wavers when presenting the reader’s work in the production of the mental model; it’s either an individual “construction” or a “reconstruction” of what was encoded by the author. However, his comparison of it with the notion of a “discourse model” in linguistics, which implies an ideal of collaboration between interlocutors as they draw inferences in a conversation, seems to indicate a preference for a match between the mental representations developed by the reader and the author. The narrative text provides a sort of blueprint for the creation of the mental model.204
The (re)construction of a storyworld involves two general tasks.205 On the one hand readers need to make an inventory of “narrative microdesigns” to understand what is going on at each moment in the text; they need to identify states, events, and actions, and they need to have an idea as to how all these elements of the storyworld can be part of larger sequences. For more detailed suggestions about what happens on the microstructural level, David Herman turns to Paul Werth’s text world theory and Catherine Emmott’s description of contextual frames. Werth makes a distinction between “world-building elements” that create a background (such as deictics and referential nouns) and “function-advancing propositions,” which are utterances containing an actor and an action involving a goal.206 Emmott describes a contextual frame as “a mental store of information about the current context, built up from the text itself and from inferences made from the text.”207
While David Herman’s suggestions as to the microdesigns certainly allow for a fine-grained description of the ways in which readers slowly (and often semiconsciously) develop their mental models of the fictional world, they do not result in an interpretive tool. But readers also have to take into account principles for “narrative macrodesigns” such as overall perspective and the general evocation of time and space. In addition, the macrodesigns of storyworlds include the principle of “contextual anchoring,” the way in which a text develops an “interface” with its interpreters.208 While many of the reader tasks on the microlevel may seem automatic, those on the macrolevel can be rather complicated, not least because over a long stretch of text the combination of various cues can allow for quite a bit of leeway when it comes to interpretation. As Herman further explains, with a nod to possible world theory, narrative sequences are not assembled by adding up all the states, events, and actions but are instead compiled “ecologically”—by measuring the results against other possible trajectories in the fictional world under construction.209 An important aspect of the entire operation is Ryan’s principle of minimal departure, according to which readers assume that there is a high degree of similarity between their own experience and the fictional world.210
In his treatment of spatialization, David Herman emphasizes that the structuralist understanding of narrative as a temporal type of discourse has led classical narratology to prioritize the study of time over that of space. Work in linguistics enables postclassical narratology to try and set the record straight. Here is a primer of the research appropriated by Herman for the analysis of the construction of space in the storyworld. Deictic shift theory suggests that all storytellers cue their audiences to forgo the parameters of the world in which they are processing a story in favor of those obtaining in the storyworld.211 In simple terms “here” and “now” start referring to the place and time in which the narrative is set. Herman does not hesitate to use a spatial metaphor for the shift in question: “Story openings prompt interpreters to take up residence (more or less comfortably) in the world being evoked by a given text.”212 The degree of comfort may depend on, for instance, the complexity of the fictional world.
Herman also turns to semantics, where spatial expressions are said to exhibit a mutual dependency between figure (a specific object) and ground (a background for the foregrounded object). In the work of Barbara Landau and Ray Jackendoff on which Herman builds, the relationship between figure and ground has led to the description of space in discourse as a region, in which the figure is located through the use of a spatial preposition, as in the expression, “The cat [figure] is sitting on the mat [ground].” Language also expresses paths or trajectories to designate a figure’s motion or orientation (along a horizontal or vertical axis).213 Furthermore, the spatial specifics of the storyworld can be evoked with reference to the opposition between invariable topological locations and projective locations that depend on the way in which they are viewed. These acts of viewing relate to the narrative macrodesign of perspective and may thus become an important aspect of an interpretation of space in a narrative text.
David Herman’s linguistic angles on the construction of space do not amount to a coherent approach to the spatial aspects of the storyworld. In one of the most elaborate treatments of narrative space to date, Katrin Dennerlein also adheres to the notion that the fictional world is a mental model. Starting from folk psychology to develop her own notion, Dennerlein first defines space generally as “a container that allows for a distinction between inside and outside. Every space may be contained in a bigger space, and any space consists of discrete smaller spaces.”214 For her conception of space, Dennerlein foregrounds the opposition between inside and outside at the heart of this relatively simple definition, and she restricts her investigation to objects and areas that constitute a potential environment for figures in the narrated world—“objects in which these figures find themselves or which they can move into.”215 Of course when it comes to literature, these objects may be fictional.
Dennerlein works within a theory of communication as a process of inferences that ultimately establish what the speaker intended, but she does not overlook the specifics of literary communication that complicate this outcome. In order to solve this paradox she uses the notion of the “model reader”216 (with “model” meaning “ideal”) to propose that space in narrative amounts to a mental model217 (with “model” meaning “a coherent whole”) on the part of such an informed agent. In order to consider the textual cues that permit the creation of the model, Dennerlein distinguishes between fictional and nonfictional texts. In the case of the latter, real-world knowledge can easily be activated by the model reader to enhance the picture of the spaces mentioned in the text. When the fictional text provides suggestions about real-world spaces, the model reader will activate the necessary knowledge, but narrative fiction can also lead to spatial inferences about non-narrated locations in other indirect ways—characters can, for instance, summon spaces related to their profession (as in the case of a butcher), and certain events and actions will have spatial implications (as when a narrator brings up an exchange between a judge and a witness).
In her efforts to describe the elements that contribute to the mental model of space in the narrated world, Dennerlein divides her attention between story and discourse. On the level of discourse, she assumes that the spatial component of a particular “situation” (her alternative term for Emmott’s contextual frame) is central to the viability of the mental model, because it will allow the reader to reimagine a situation when necessary. Building on the cognitive notion of an “object region” (the range in which people use a certain object), Dennerlein proposes the “event region” to cover the fact that narrative fiction can create its own unusual versions of such a space. An event region becomes a “setting” (Schauplatz) when it amounts to the deictic center of the narration, but that is not always the case.
The text type can be decisive when analyzing the narrator’s management of spatial information, and Dennerlein illustrates this by defining a description as a text type that conveys the stable aspect of, for example, a town square without any reference to a specific event. On the level of story, Dennerlein considers the influence on narrated space of what she calls “spatial models” (Raummodelle)—collections of knowledge about action sequences and the spaces they include. The institutional spatial model “prison” contains not only a set of distinct spaces such as the gate, the cells, and the recreation area but also typical actions such as the escape attempt and the rounds by a guard.218 Dennerlein also brings in generic action schemes to investigate to what extent an individual text handles the spatial choices and emphases contained in such a scheme. Does a contemporary spy novel, for instance, rehearse the genre’s regular preference for broad descriptions of exotic locales, or does it concentrate on the spaces of the story events themselves? When it comes to the physical characteristics of narrated space, Dennerlein falls back on her definition of space as an environment for human figures to emphasize the centrality of the classical axes of experience (up/down, left/right, and forward/backward), which lead to the prominent indication of relative positions.
The storyworld of Wasco’s “City” is strongly determined by the first panel after the title, in which a little spaceship seems to have just landed in an urban environment. Since the protagonist and his or her dog seem to have come from afar (perhaps even from another planet), they are quickly understood to be exploring what to them (and by extension the reader) might well be a new locale. The fact that they fly away at the end reinforces the idea that they might have come on a reconnaissance mission or at least that they are interested in the city. While the action scheme of a city visit from outer space (to which the opening and closing panels appeal) is relatively simple and familiar, the city itself is so out of the ordinary that its strangeness comes to dominate the storyworld.
This is a weird place indeed. Panel one already has two special walls and an entrance without an actual door.219 Panel two shows a sidewalk that crosses the street, probably making it hard for cars to pass by. Maybe these streets are not meant for cars at all; in panel twelve the pointed obstacles on the sidewalk oblige the visitors to walk in the “street,” where there are also big works of art, and in panel seventeen there is a bench in the street instead of on the sidewalk. Panel two also features two uncovered holes in the sidewalks, and part of a building seems to have been carved out to accommodate one of them. Holes like this return in panels four, eight, fifteen, and seventeen, and in panel nineteen the end of what could be a spire even protrudes from one of them. What are they for? A sequence of holes occupies the place for a crosswalk in panel four, and in panel seventeen they are also placed in the middle of a “street.” Do they constitute an intentional hazard, or is this reaction on our part the consideration of readers who are stuck in the functionality of a normal city and motivate an aspect of their storyworld for “City” accordingly?
Another fascinating characteristic of Wasco’s city is its system of walkways. First seen in panel three as a flat passage above the street from one building to another (and clearly not meant for a cold climate), they become steeper and more challenging as the boy or girl and the dog move along; in panel eleven the walkway even looks like a slide. There are also strangely pointed buildings, bizarre sculptures, entire buildings on roofs, and an electric chair on top of a tower (panel ten) but hardly any signs of life. If what stands behind the wall in panel fifteen are trees or tall plants, it looks like they are dead, and the same is true of the vegetation on the space next to the wall in panel sixteen.
The emptiness of the town already becomes evident in panels two and three, from which the visitors and possible inhabitants are absent (but of course these panels could be what the visitors see at the outset of their walk). Still, in panel six the protagonist points to a bird perched high up on a building, and in the next panel a bird is flying in the air. Moreover, in panel fourteen there is something brown flowing into what could be an open sewer, which might suggest life in the city is simply hidden from view. Perhaps the visitors have landed at the time of the siesta and maybe the whole town will soon awaken. Many roofs sport various kinds of antennas, which might capture data for humans, and the panorama in the wide bottom panel has five smokestacks, which also suggest some form of human activity. The small plant in a window on the left side of the panorama panel could be dead, but that is not entirely clear.
What is more important is that the city is clean, which the reader may gradually understand as a decisive confirmation of human presence. In that case the strangest aspect of the storyworld is the fact that the path of the visitors through the city remains unperturbed. The visit from outer space that is part of our cultural repertoire typically contains some form of conflict. Otherness and normality are bound to clash. Here, however, the visitors have all the time in the world to take everything in, and in panel seventeen they are even seen to rest on a bench before returning to their spaceship. So, what can readers make of this calm in the strange storyworld they are bound to construct? Is it deceptive—did the information coming in through the antennas warn the local population about the visit, so that they could hide inside? The sophistication of their architecture and their art certainly doesn’t rule out this possibility, and the joint appearance of Christian and Muslim symbols (the cross and the crescent moon) on top of the wall in panel fifteen and of the building for worship on one of the roofs in the final panorama also suggests that they have a knack for keeping the peace. But what about that wired chair in panel ten? With its connotation of death, doesn’t it confirm the treacherousness of the holes all over the city, holes in which people can easily disappear? Or has the entire population of the city perhaps vanished into these holes, to crawl out again only when the strangers have disappeared?
A graphic narrative such as “City” suggests that the notion of storyworld is useful precisely because it highlights the gaps a reader has to fill in order to process a story in a more or less coherent and satisfying way. The mental model of the city in “City” derives from the interaction of event and environment as it activates the reader’s knowledge. Contrary to Dennerlein, we have shown that Wasco’s text makes it difficult to describe the model reader she has envisaged. The questions in our interpretation above cannot get definitive answers. While Dennerlein’s proposals for a narratology of space are more elaborate and precise than David Herman’s suggestion about spatialization and the construction of the storyworld, his emphasis on the individual reader’s contribution seems more appropriate for a theory of narrative that wants to incorporate the mental models involved in the act of reading.
Telling stories is a form of communication. In classical narratology the study of narrative communication was restricted to the interaction between fictional agents such as the narrator, the character, and the narratee. Reference to the real author and reader was considered unscientific, since science deals with abstract and universal structures, not with concrete and contextualized agents. Postclassical narratology has a different view on science and has turned its attention to the author and the reader and their contexts in an effort to grasp the message of the text. Roman Jakobson’s communicative scheme, linking sender and receiver by means of the text and the context, is never far away in these approaches.220
More generally, two frames of reference appear time and again: rhetorics and pragmatics. Rhetorics considers a story as an attempt to persuade the reader by means of all kinds of devices. These devices themselves are no longer analyzed in their own right—as in structuralism—but they are studied in terms of the speaker’s intention, as well as their orientation to and effects on the readers. Insecure narrators may have different intentions: perhaps they want to make their readers insecure as well, or they may want to seduce the readers or make them curious. The nature, meaning, and function of a narrative strategy become clear only when these effects are taken into account. It is no longer sufficient to limit oneself, as a structuralist would, to the relationship between the narrator and the fictional universe.
Ross Chambers, for instance, sees the interaction between text and reader as a form of seduction. Narrative techniques aim to seduce the reader, who adapts these techniques to his or her own desires. Texts become readable only by the transaction between seduction and desire, a process in which narrative strategies and characterization play a decisive role. In this transaction the text attains its value, and the reader assumes his or her responsibility with regard to the text by responding appropriately to the seduction strategies. What amounts to an appropriate response is partly thematized in the story, which indicates through characters and narrators what the good listener or reader looks like. Chambers considers, for instance, the character of Félicité from Flaubert’s Un cœur simple as a role model for the reader.221 But the “appropriate” reaction is partly determined as well by what readers themselves find adequate. However, in this approach the reader is never absolutely free. He or she has to take into consideration not only the text but also the intention of its author. This leads to pragmatics, which studies a text as a form of communication, with a sender, a message, and a receiver. The interaction between these three communicative agents is the core business of rhetorical narratology.
James Phelan and Peter Rabinowitz, undoubtedly the most influential proponents of rhetorical narratology, define narrative “as a purposive communication of a certain kind from one person (or group of persons) to one or more others. More specifically, our default starting point is the following skeletal definition: Narrative is somebody telling somebody else, on some occasion, and for some purposes, that something happened to someone or something.”222 This definition clearly builds on the pragmatic and rhetorical traditions we have just mentioned. To understand narrative from this perspective, one must study the “feedback loop among authorial agency, textual phenomena (including intertextual relations), and reader response. In other words, our approach assumes that texts are designed by authors (consciously or not) to affect readers in particular ways[,] that those authorial designs are conveyed through the occasions, words, techniques, structures, forms, and dialogic relations of texts as well as the genres and conventions readers use to understand them.”223
For rhetorical narratology, the balancing of sender, text, and reader is fundamental to the production and interpretation of narrative. None of these three agents should be studied in isolation. The text does not constitute an autonomous world, but it is a message that contains traces of the author’s intentions and signposts for the benefit of the reader. Various concepts have been proposed to make the critic’s balancing act more concrete. The most famous one is undoubtedly the “implied author,” which we discussed in chapter 1. Phelan defines the implied author as “the consciousness responsible for the choices that create the narrative text as ‘these words in this order’ and that imbues the text with his or her values. One important activity of rhetorical reading is constructing a sense of the implied author.”224
This construction should be in line with the intentions of the (implied) author. If that is the case, the reader is part of what Rabinowitz calls the “authorial audience,” a hypothetical and ideal audience that the author had in mind while composing the text: “Authors are forced to guess; they design their books rhetorically for some more or less specific hypothetical audience, which I call the authorial audience. Artistic choices are based upon these assumptions—conscious or unconscious—about readers, and to a certain extent, artistic success depends upon their shrewdness, on the degree to which actual and authorial audience overlap.”225 The same goes for the success of the reader’s interpretation: a reading is successful only if the actual audience nearly coincides with the authorial audience.
Clearly, rhetorical narratology is normative: it proposes norms for the successful telling and interpreting of stories. The authorial intention may be the central normative aspect, but it is itself not beyond criticism. Authors may fail to express their intention successfully. More generally, intention is the beginning of creation and interpretation, but it needs to be evaluated too: “once you decide to take a rhetorical perspective, the best way to make initial sense of texts is to treat them as if they are intended to be made sense of—and then, once we’ve reconstructed that multidimensional sense, we can take the next step of evaluating the author’s communication.”226 For instance, Phelan and Rabinowitz “greatly admire” in some passages of The Adventures of Huckleberry Finn “Twain’s handling of the relationships among author, narrator, and audience,” whereas they criticize “Twain’s far less successful narration” in other places.227
This normative and balancing approach goes hand in hand with the idea of narrative as a pact between sender and receiver. Rabinowitz describes this contract in terms of rules the reader has to follow in order to re-create authorial intentions. First, the rules of notice: a reader pays attention only to certain aspects of the text; others are often simply ignored. Second, there are the rules of signification, which are used by the reader to assign a (possibly symbolic) meaning to the aspects that attract his or her attention. This consists of connecting these aspects to the reader’s everyday experience by interpreting, for example, characters as if they were actual human beings with a specific psychological profile. Third, the reader uses the rules of configuration to connect different textual fragments to each other. This creates patterns that are neither exclusively textual nor exclusively determined by the reader’s expectations but are instead the results of a fusion between the two. Finally, the reader applies the rules of coherence to transform the text into a coherent whole that nevertheless leaves room for paradoxes and deviations. These rules function as a kind of contract between author and reader. If they follow different rules, the creation and/or interpretation cannot be successful.228
To apply these rules, readers have to look for textual traces left by the implied author and aimed at the authorial audience. Rhetorical narratology has not only proposed concepts relating to the sender (e.g., “implied author”) and the receiver (e.g., “authorial audience”) but has also specified the way the text can steer the communication between its sender and receiver. Most famously, Phelan has distinguished between mimetic, thematic, and synthetic aspects of the narrative. These aspects can be found on the level of narrative components, such as character and setting, but they also imply a specific attitude toward them on the part of the reader.
As to the first level, Phelan describes the three components of character as “the mimetic (character as person), the thematic (character as idea), and the synthetic (character as artificial construct).”229 The mimetic component is defined as “that component of character directed to its imitation of a possible person. It also refers to that component of fictional narrative concerned with imitating the world beyond the fiction, what we typically call reality.”230 The thematic dimension designates “that component of character directed to its representative or ideational function; more generally, that component of a narrative text concerned with making statements, taking ideological positions, teaching readers truths.”231 The synthetic, finally, refers to “that component of character directed to its role as artificial construct in the larger construction of the text; more generally, the constructedness of a text as an object.”232
These dimensions imply a particular stance on the part of the reader. The reader focusing on the mimetic aspect accepts the fictional world as real. He or she is part of what Rabinowitz and Phelan call the “narrative audience.”233 The authorial audience, on the other hand, focuses on the synthetic dimension: it zooms in on the construction of the narrative world and on the principles governing it, that is, it aligns itself with the authorial project and intention: “The authorial audience takes on the beliefs and knowledge that the author assumes it has, including the knowledge that it is reading a constructed text.”234 Readers paying attention to the thematic component tend to go for the ideas and ideology of the text.
Of course the different attitudes usually go together. The “actual audience,” that is, the empirical readers, embraces both the authorial and the narrative audience. It is a matter of stress and degree, not of exclusivity. The authorial audience seems to have the lead here, as its respect for the authorial construction necessarily implies an awareness not just of the synthetic but also of the mimetic and thematic dimensions.235 Compared to the reader who is immersed in the narrative, the authorial audience has a richer consciousness: “the authorial audience has the double consciousness of the mimetic and the synthetic, while the narrative audience has a single consciousness.”236 In every reading there is a continuous and changing interaction between the three components, generating a dynamic and a tension that are seen as the essence of reading.
While narratives are studied as a continuous exchange between sender, text, and receiver, dynamics are at the heart of all rhetorical readings. Unreliable narration, for instance, is not reduced to one final source (the implied author) but studied as a continually changing interaction between the three “primary tasks” of narrators: “they report (along the axis of facts, characters, and events), interpret (along the axis of knowledge or perception), and evaluate (along the axis of ethics).”237 The authorial audience evaluates these activities and tries to find out whether the narrator is concealing information (underreporting) or distorting it (misreporting). This leads to six types of unreliability: underreporting, underinterpreting, underevaluating, misreporting, misinterpreting, and misevaluating. Narrators may combine these types or shift from one category to another. It is up to the reader to keep track of these dynamics and this polyphony.238
Following M. M. Bakhtin, Phelan does not consider a narrative text as a single-voice monologue that supposedly addresses the reader in a compelling manner but rather as an exchange of voices in which the reader has an active role in weighing one voice against another.239 When reading a story, a reader hears the voices of all kinds of narrative agents—both inside and outside the story—and tries to distill from this polyphony one harmonious whole. This is precisely the way in which the reader gets actively involved in the story. This involvement links “the logic of the text’s movement from beginning to middle through ending (what we call textual dynamics) and the audience’s temporal experience (readerly dynamics) of that movement.”240 If these two dynamics run parallel, they lead to (or at least give a good sense of) the authorial intention.
It is not just the level of narrating and telling that leads to a dynamic reading experience; the level of the told (i.e., the sequence of events) is witness to the same kind of interaction between “textual and readerly dynamics.”241 In rhetorical narratology the plot is no longer reduced to a structuralist scheme of possibilities; it resides in the interplay between textual and readerly dynamics, which point to the authorial project. A plot “typically proceeds through the introduction, complication, and resolution (in whole or in part) of unstable situations within, between, or among the characters. These dynamics of instability may be accompanied by a dynamics of tension in the telling—unstable relations among authors, narrators, and audiences—and the interaction of the two sets of dynamics, as in narratives that employ unreliable narration, may have significant consequences for our understanding of the ‘something that happened.’”242
Since the story of “The Map” begins in a bookshop that offers the public exactly what it expects and wants, one may take this setting as an (authorial) indication of the synthetic and metafictional nature of the story. The scene in the shop seems to suggest a general view on fiction as a means for entertainment, tailored to suit the expectations and tastes of the readers. Literature is part of a provincial kind of coziness and contentment. In the second part of the story, where the seemingly omniscient and impersonal narrator turns out to be homodiegetic, this general idea is particularized via the main character’s desire to have his small-town reality mapped, that is, “translated” into the map he bought at the bookshop. This desire seems to tie in with the mimetic attitude. But there is more to it. The relation between the map and reality works both ways: it is not only exciting to see the real world mapped, it is equally thrilling to visit the parts that are on the map and that the narrator has never seen before. The map is not just a representation of reality; reality must adapt itself to the map. This two-sided form of dynamics can be read as a (synthetic) indication of the interchange between the textual dynamics (mapping the world) and the readerly dynamics (the reader “performing” the script or following the route laid out by the text).
However, when the narrator begins to realize that a complete overlap between map and route (or text and reality) is not possible, he loses interest. This may suggest that mimetic readings sooner or later meet their limits and that it is necessary to move beyond mimetic desires. The story then is not just about living and growing up in a provincial town; it is about becoming a writer by moving beyond the diverting and mimetic functions of texts. It speaks about the nature and function of fiction, as well as about its relation to reality. The tensions and interactions between map and route, text and reality, can also be seen at the level of narration, where the seemingly omniscient narrator whose mapping of a world is to be accepted as a reality makes room for an I-narrator who is not omniscient and who does not succeed in reconciling map and route, text and reality. If the first paragraph gives the safe and comfortable impression of a realistic story about small-town life, the shift in the second paragraph leads to an uneasy feeling that the narrator is not in control of his story and that his fictional world fails to overlap with the real one.
The issue of fictionality has gained prominence in recent rhetorical narratology, which rejects classical, intratextual approaches to fiction. In The Distinction of Fiction (1999), Dorrit Cohn tried to pinpoint the uniqueness of fiction by means of textual features. In the seventh chapter, “Signposts of Fictionality,” she offered three narratological features to distinguish fictional from nonfictional texts. First, fictional narratives contain two levels of analysis, story and discourse, whereas nonfictional texts (such as a historical study) feature a referential level. Second, only fiction offers complete access to the minds of others. There can be no omniscience outside fiction, meaning “that the imaginary figures can be known in ways that those of real persons can not.”243 Finally, there is “the duplicate vocal origin of fiction.”244 It refers to the author/narrator distinction not found outside fiction.
There have been many other attempts to find “signposts” of fictionality.245 Cohn herself referred to Käte Hamburger’s The Logic of Literature. Hamburger claimed that “epic fiction is the sole epistemological instance where the I-originarity (or subjectivity) of a third person qua third person can be portrayed.”246 Seeing a third person as if it were a first person amounts to the second distinction of fiction Cohn mentioned. Free indirect speech is an exemplary technique for merging the voices of the third and first person, and as such it is typical of fictional texts.247 Other linguistic and technical devices have been proposed as signposts of fictionality, for instance by Jean-Marie Schaeffer, who pointed to certain types of verbs, deictics, and dialogues.248
Numerous objections have been formulated against these intratextual approaches. None of the so-called “signposts” are exclusive to fiction, and fictional texts often use devices that are supposed to be signposts of nonfiction, such as controllable references to the real world. Boundaries between fiction and nonfiction seem to be porous and unstable; instead of talking about a dualism, it seems better to speak of a continuum with varying degrees of fictionality. To come to terms with these and similar problems, rhetorical and pragmatic approaches do not regard fictionality as a distinctive textual feature of fiction-as-a-genre but as a communicative and discursive strategy. In The Rhetoric of Fictionality (2007), Richard Walsh claims that fictionality “functions within a communicative framework: it resides in a way of using a language, and its distinctiveness consists in the recognizably distinct rhetorical set invoked by that use.”249
The distinctive set is no longer a collection of textual traits. It is part of a strategy the reader uses to make sense of the communication in which he or she is engaged. Relevance, rather than truth or referentiality, is the decisive factor here: if it is relevant to read the text, or parts of it, as fiction, the reader will do so. In that way he or she will make the most of the message and will get close to the intentions of its sender, or so the reader imagines. The three parties typical of the rhetorical approach (sender, message, receiver) are united by the assumption of relevance in communication: “Fictionality is neither a boundary between worlds,250 nor a frame dissociating the author from the discourse, but a contextual assumption by the reader, prompted by the manifest information that the authorial discourse is offered as fiction. This contextual assumption is a preliminary move in the reader’s effort to maximize relevance.”251
As a consequence, fictionality is not confined to the realm or genre of fiction. Nor does it coincide with narrative. There are fictional and nonfictional narratives, and fiction can be used in non-narrative environments, such as poetry. More generally, fiction is a mode that can be used, both by the sender and the receiver, in serious, truthful, and referential forms of communication to maximize relevance. In their path-breaking essay “Ten Theses about Fictionality,” Henrik Skov Nielsen, James Phelan, and Richard Walsh point to the ubiquity of fictionality “in politics, business, medicine, sports, and throughout the disciplines of the academy.” It is “a specific communicative strategy” and an important “vehicle for negotiating values, weighing options, and informing beliefs and opinions,” but it can function only if a tacit understanding is shared by sender and receiver. The speaker and the audience must “share an understanding of the distinction between fictionality and nonfictionality.”252
All rhetorical narratologists agree that fictionality is used and should be studied in the interplay between sender, text, and receiver, but different theorists may emphasize different parties in that interplay. To Phelan, the decisive role in this process of understanding and negotiation seems to go to the sender and his or her intention. For him, fictionality refers “to any rhetorical act in which somebody on some occasion intentionally signals his or her use of a discursive invention to someone else for some purpose(s).” He adds that “the definition: someone intentionally signaling distinguishes fictionality from lying, on the one hand, and from unintended inventions such as dreams on the other.”253 So, the distinction of fiction seems to be primarily an intentional and communicational matter.
To Walsh, the reader, though by no means free, seems to be decisive, not only because he defines fictionality as “a contextual assumption by the reader,”254 or because he recognizes that communicational intentions are inferred,255 but especially because he goes to great lengths to show that the sender is never in complete control of the communication. Walsh explicitly states, “I would reject the conclusion that the rhetoric of fictionality must therefore be wholly accounted to authorial intention, not least because any such model of novelistic communication is necessarily abstracted from the particularity of the narrative, which would therefore be underdetermined.”256 Stefan Iversen and Henrik Skov Nielsen focus on the self-conscious dimension of the message as the crucial feature of fictionality, stating that “fictionality is present whenever the invented nature of a communication is signaled.”257 Obviously that does not rule out authorial intention or readerly inference;258 it is merely a matter of emphasis. Whether the stress is on the sender, the receiver, or the message is not a fundamental or a divisive issue, because the communicative outlook remains the same in the works of Phelan, Walsh, Nielsen, and Iversen. Furthermore, they all agree that fiction is not an autonomous world or a self-contained genre but that it is geared toward reality. This “double exposure of imagined and real” recurs in another important communicative kind of narratology, inspired by cognitive studies.259
A narratological interpretation differs from reader to reader, and the most progressive type of narratology is the one that takes into account the interpretive variants in the theory formation. In structuralist narratology the reader was officially excluded, but whenever the abstract categories and types were actually applied, their effect on the reader was implicitly referred to. Delay and acceleration, omniscience and reliability, can be grasped only as effects on the reader. This is no coincidence: the structuralists may have isolated the text, but they still worked with the well-known communication triad of sender, message, and receiver, in which the audience obviously plays a central role.
From the end of the 1960s onward, Wolfgang Iser and Hans Robert Jauss started to develop, respectively, reception aesthetics and reception history—two theories that for the first time included the reader in literary criticism in a systematic way.260 Jauss wanted to rewrite literary history by considering it as a sequence of ways in which a literary text confirms and breaks with patterns of expectation. Iser, on the other hand, wanted to design a theory describing the effect of the literary text in general. He started from the assumption that the text pushes the reader in the direction of a certain interpretation. In analogy to Booth’s implied author, Iser opted for the somewhat misleading term “implied reader” for the text-driven role of the reader. According to Iser, the turning points in the development of this role are the “gaps” that result from the fact that the text can never fully meet the expectations that readers cherish based on their experience of reality.261 Gaps arise not only from the imperfect overlap with reality as it is experienced but also from the structure of the text. They can, for instance, result from abrupt transitions in the story or sudden changes in narration or focalization.
Iser’s theory is discussed in the last chapter of Rimmon-Kenan’s survey Narrative Fiction for a particular reason. She realized all too well that narratology had to take the reader into account more systematically, and at the time she was writing her textbook, Iser’s work provided a natural supplement in this respect. Rimmon-Kenan also paid attention to other reader-oriented approaches such as psychology and semiotics.262 In the afterword she added in 2002 to her now classic book, she emphasizes the importance of these approaches even more. She appreciates the expansion brought about by postclassical narratology, which no longer limits analysis to literature itself but also pays attention to the context. According to her, this expansion is largely due to various reader-oriented approaches.263
Ever since reception aesthetics, the reader has been integrated into narratology in many different ways, especially in the context of the cognitive paradigm, which has acquired an important position in the human sciences and which stresses the processing of information.264 The terms “cognitive” and “information” may create the impression that elements of affect do not play a significant role in this approach, but this is not necessarily the case. Psychologists such as Richard Gerrig and Victor Nell study the emotional identification and far-reaching immersion that give readers the impression they are being carried away by the text.265 In their empirical tests researchers working within this paradigm mostly limit themselves to the concrete interpretations of (parts of) texts.
In a very extensive account of research on language processing and narrative comprehension, Sanford and Emmott aim to integrate contributions to these topics in neuroscience, psychology, and psycholinguistics with work done in the humanities. They propose a “rhetorical processing framework” with three strands. “Fundamental scenario-mapping theory” develops the basic assumption that “much understanding is accomplished by relating what is being said to background knowledge,”266 or, as they also call it, “everyday world knowledge.”267 The “rhetorical focusing principle” considers the ways in which the reader’s attention is manipulated by the author, and as such it amounts to a “psychological version of the humanities ideas of foregrounding and defamiliarization.”268 The third strand of the processing framework deals with effects of reading (such as immersion in the storyworld) as they are approached empirically in studies of embodiment in psychology.
Even within the discipline of narratology, some cognitive approaches come very close to cognitive psychology and adopt some of its positivistic presuppositions. Peter Dixon and Marisa Bortolussi start from a rigid distinction between, on the one hand, so-called objective characteristics in the text and, on the other hand, subjective effects of those characteristics on the reader.269 As we will also see in our discussions of narrative ethics and postmodern narratology, most postclassical approaches reject this traditional and dualistic view of the objective text and the subjective reader. So-called objective characteristics are construed by the reader as well. The exact points of contact and the differences between text and reader can probably not be determined in any straightforward way. For those approaches, a rigid distinction between object and subject is certainly no solution.
Since this handbook is geared to the relevance of narrative theory for the interpretation of literary texts, our presentation of cognitive narratology will be narrowed down to a discussion of how, as David Herman puts it, “stories across media interlock with interpreters’ mental states and processes, thus giving rise to narrative experiences.”270 This question “bears on stories viewed as a target of interpretation; it concerns ways in which interpreters use various kinds of semiotic affordances to engage with narrative worlds.”271 In this section, therefore, we engage with borrowings from the cognitive sciences by narratologists in order to see how they can enhance the meanings of our three central stories. This means, for instance, that when we discuss how readers construct characters, we will zoom in on the contributions by Ralf Schneider,272 since they go beyond an abstract model of this process so as to describe exactly how the construction comes about.273 With reference to the reader’s engagement with a narrative and the emotions that can produce, we will draw special attention to the work of María-Ángeles Martínez, whose notion of “storyworld possible selves” also allows for practicable analysis.274
A cognitive approach that has proven useful for literary theory is frame theory. This approach was developed by, among others, Marvin Minsky in the study of artificial intelligence.275 It assumes that people, when confronted with a new situation, will select a structure from memory to help them to deal with this new situation. For Minsky this frame consists of a network of nodes and relations that can be applied quite easily to the new situation. The theory also allows for some flexibility in this application: on the one hand the network consists of representations that are perfectly valid for the situation; on the other hand, it also contains a number of slots to be filled with specific details that are valid only for the situation at hand and thereby augment the frame’s relevance for that situation. These details can be worked out to fit the new situation in a concrete way. When in a foreign country one notices a sign on the edge of the road, its location, shape, size, and the material it is made of will immediately trigger the insight that this might be a traffic sign. For the precise meaning of the sign, the icon and text on the sign will be as important as the concrete environment.
Manfred Jahn defines a frame as “the cognitive model that is selected and used (and sometimes discarded) in the process of reading a narrative.”276 Both Jahn and Ansgar Nünning have applied frame theory to narratology.277 As we mentioned in the discussion of the structuralist characteristics of narration, Nünning argues that the unreliable narrator is often a psychological projection of the reader who aims to clarify ambiguities or contradictions in the narrator’s utterances. This argument, however, does not capture all meanings of the concept. It is necessary to anticipate a whole gamut of reader reactions. In any case, textual features are not sufficient to argue against the trustworthiness of a narrator. One always has to investigate the framework leading to the observation that the narrator is lying or is morally unstable.
Narratology traditionally reserves the term “lie” for a lack of correspondence between a narrator’s utterance and a situation that occurs in another part of the text. If the reader accepts the situation as an actual one, the utterance is called unreliable. At first glance, this is an intersubjectively valid method for distinguishing truth from falsity, and yet even this method must take into account cultural and epistemological nuance. What about “white” lies and formulations that can be interpreted in different ways? Judgments concerning moral instability imply a general norm that, at least in our society, can hardly be defined. Every interpretation of a narrator’s unreliability has to be linked to the specific norm one uses as a reader.
The first-person narrator in Gerrit Krol’s story misrepresents the period before he acquires the map because of the influence the map has had on him. He was liberated by it, and therefore he is no longer capable of representing the time before the liberation in any objective way. In other words, the first-person narrator in “The Map” is not entirely reliable when he is talking about his youth. He does not lie, but his selection of data gives rise to a biased image. In the construction of that image the reader plays a role as well. He or she may interpret the first paragraph ideologically as the description of an unfree world dominated by the Christian worldview. Sunday is the day of the Lord, and people are supposed to rest. This frustrates a number of desires related to consumption, which nicely illustrates the hold of Christian ideology on society. If no connection is made between the shaded shop windows and Christian morality, the entire first paragraph is much less likely to be interpreted ideologically. As a result, the paragraph will appear less of a distortion originating from the narrator’s evolution, an evolution that may be described as a distancing from or even a rejection of the original ideology. Perhaps these descriptions are in fact too strong, and in any case they make sense only within the ideological cognitive frame some readers use.
Manfred Jahn uses frame theory for the analysis of third-person narrative situations. Following Mieke Bal, he reduces narrative texts to the formula “X tells R that Y sees that Z does something.”278 According to Jahn, this formula is the most complete description of the model a reader can use to process a third-person narrative text. There are three typical models, the so-called defaults. First, there are texts for which the entire formula holds. Second, there are texts without an internal focalizer.279 These texts correspond to the formula “X tells R that Z does something.” And third, there are texts with a nearly invisible narrator: “(X tells R that) Y sees that Z does something.” Every model comes with its own expectations. A visible narrator who does not participate in the story is believed to be omnipresent, omniscient, and reliable. From an internal focalizer we expect an especially subjective presentation of the events that could, as it were, be seen to originate in the focalizer’s mind.
As we have indicated in our brief presentation of the theory, the concept of frame is flexible enough to adapt defaults to texts that do not immediately conform to the typical expectations. In the course of reading, the selected frame is developed further and refined at the level of the details that have to be filled out.280 If the text completely clashes with one’s basic expectations—for instance, when a first-person narrator disrupts the third-person model—the initial frame is dropped. The choice for a new frame depends not only on the solution it can offer to the reading problem but also on the extent to which it provides a better interpretation of the text read so far.
Jahn argues, however, that some narratological terms and concepts have to be made more flexible if they are to be used as efficiently as possible in the processing of narrative texts. A sentence such as “The room was dark” may lead to the application of the well-known frame “description,” a mode of narration Seymour Chatman describes as stopping the narrative clock so as to let the narrator present the environment.281 However, this frame is too narrow for this sentence, since it may also be the result of an internal focalizer’s observation. In that circumstance, time does not stop at all: the character’s observation implies a certain duration.
Another concept that, according to Jahn, has to be made more flexible is that of free indirect speech. His frame version of free indirect speech emphasizes the context in which this form of representation occurs. A sentence recognized as free indirect speech (frame A) can be part of a character’s observation (frame B), and frame A as well as frame B may fit into a quotation of the character’s thoughts by the narrator or into a summary of those thoughts also provided by the narrator. An example from Madame Bovary: “Would she never escape? She was every bit as good as all the women who lived happy lives.”282 Frame A consists of the sentence “Would she never escape?”; frame B is Emma Bovary’s observation (she asks herself this question); the narrator quotes her thoughts. In the second sentence (“She was every bit as good as all the women who lived happy lives”), it is very well possible that the narrator is summarizing her thought.
In order for a sentence to be recognized as free indirect speech, Jahn develops a general description listing three characteristics. First, a sentence that can be read as free indirect speech is a nonsubordinated construction: it is not a subclause introduced by “that” but an independent main clause. Second, tense and person are adapted to the existing narrative situation. “Would he see her tomorrow?” is the free indirect version of “He wondered, ‘Will I see her tomorrow?’” in which the first person shifts to a third person and the future tense changes to the conditional mood. Third, the sentence represents thoughts, utterances, and writings of a character. Because of this general description, the literary standard model of free indirect speech (third person, past tense) loses its focal status in narratology. Nevertheless, Jahn does not go as far as Fludernik, who as we mentioned at the end of chapter 2 thinks that free indirect speech should be studied as a form of typification.283
“Pegasian” is a third-person narrative with a reclusive, invisible narrator, which makes the reader’s task slightly more difficult. The riding school itself does not become the subject of a description and neither does the appearance of the riding master and the girl. Only the girl’s denim apparel is mentioned. This scanty information slightly frustrates the reader’s elementary expectations. Nevertheless, it can be assumed that the story is told by a traditional narrator, which means that this is an example of the first formula, and as such it answers to the most complete formulation of the framework: X tells R that Y sees that Z does something. Once this is accepted, the narrator’s invisibility and parsimony can be seen as an indication of the distance maintained with regard to the characters. The invisible narrator presents the conversation between the riding master and the girl in a mostly indirect way, avoiding a simple dialogue, perhaps to add an ironic note to the whole event. This irony may lend the girl’s opinions a pathetic overtone that is further intensified by the triviality of the topic: a pair of riding breeches.
Moreover, the indirect presentation in this text allows for a mixture of utterances and thoughts that undermines the riding master’s authority. It is not always clear whether the riding master has actually voiced a certain thought, but the accumulation of nagging opinions demonstrates the extent to which the narrator has a hold on the character and is able to humiliate him (or her). A flexible and contextual understanding of free indirect speech, as Jahn proposes it, attaches more importance to the function of this mode of presentation than to its grammatical properties. It clarifies the uses of free indirect speech and demonstrates how it ties in with the meaning of the story and the central theme of authority.
Not only does free indirect speech generate the detached depiction of the riding master, but the girl too is represented rather ironically in the first part of the story, especially when her behavior is considered to be a teenage whim. In the last paragraph, however, free indirect speech causes empathy rather than irony. The narrator recedes in favor of the character, which makes it slightly harder to use the third-person frame. The girl’s eventual insight is honored with a positive thought (“As long as you take off”) in which the narrator and the girl seem to come to an agreement.
In this interpretation the reader encounters two versions of free indirect speech in “Pegasian”: an ironic version and an empathic version. It is not necessary to drop one in favor of the other. The specific succession of distance and empathy nicely ties in with the girl’s development in relation to the riding master: at first, she completely disagrees with the instructor’s views, but this changes later. Since free indirect speech follows the girl’s development, the narrator—who is responsible for the choice of speech—could be said to be more sympathetic to the girl, in spite of relative detachment in the first part. The narrator’s ironic treatment of the riding master might suggest a shared negative attitude toward that authority figure.
David Herman published a significant share of his cognitive contributions to narratology in the flagship American journal PMLA, which underscores the prestige of this approach.284 Herman wants to know how to define a narrative text. He uses the term script, which, like frame, derives from artificial intelligence. A script is an expectation concerning the specific sequence of a series of events.285 Both in the case of frame and script, readers draw on their memory to interpret the reading experience by means of structures acquired earlier. However, contrary to frame, which is a static structure, script emphasizes development. This makes it more suitable as a theoretical tool for the interpretation of a narrative text, which is always dynamic.
According to David Herman’s general hypothesis, a story, much like a greeting or a quarrel, is a way of joining existing knowledge to new data. This can work only if we really possess the existing knowledge in question. What knowledge leads a reader to view a text as a narrative text? Herman starts from the structuralist suggestions concerning the minimal story as a temporal and causal sequence of events. He goes on to claim there are many textual characteristics that contribute to the reception of a text as a story—such as the indication of new information in a recognizable context and the suggestion of an action structure. Nevertheless, a text really becomes a story only when the reader sees a connection between the text and an existing script. The sentences “Mary was invited to Jack’s party. She wondered if he would like a kite” can easily be taken as a story or as part of a story because we all know how birthday parties are organized and prepared.286 In this example the preparation is crucial.
A specific sequence of sentences seems to constrain the number and type of activated scripts, but it is hard to deduce the exact nature of this constraint from the characteristics of the sentences themselves. In another context and for other readers the sentences about the party can be part of a story about a retirement party for a colleague who uses kites to study the weather. This context can be imagined on the basis of the text, but the full range of potential scripts can never be exhausted. It does seem feasible to say of a certain sequence of sentences that they resemble a story more than another sequence might. Herman therefore defines a text’s narrativity as the extent to which this text activates scripts urging the reader to consider the text to be a story.287 According to him, the degree of activation is higher when the activated knowledge is more complex and comprehensive. The more scripts a text activates and the more refined these scripts are, the faster the text will be considered a story.
Herman links the activation of scripts to a hypothesis about literary history. He argues that narrative innovation often implies the explicit rejection of old scripts, which forces the reader to use another kind of world knowledge in the interpretation of a text with a seemingly familiar subject. The subject is familiar, but the scripts that are normally used to interpret it do not seem to function anymore. At the beginning of Don Quixote, Cervantes makes the reader give up his or her idealized knowledge of the development of a quest in favor of what Herman describes as “scripts grounded in an awareness of human potential and limitations.”288
According to Herman, the diachronic study of reader activation has to be linked to a synchronic analysis of the various ways in which different genres deal with scripts in the same period. He gives three examples from the 1920s and 1930s: children’s literature, autobiography, and the experimental novel. In that period children’s literature also sought to activate morally virtuous scripts such as the postponement of satisfaction. Of course this presupposes a target audience that recognizes these scripts. In this context, activation implies the consolidation of existing scripts or perhaps the production of larger action structures by the combination of scripts with which young readers are already familiar.
A passage from the autobiography of Maud Gonne, an Irish nationalist, demonstrates that this genre places much higher demands on the reader, especially when it comes to scripts about identity and self. Gonne activates but also undermines the scripts in which heroism is related to masculinity, which forces the reader to revise his or her preconceived ideas about female development. Herman takes his last example from Nightwood, an experimental novel by Djuna Barnes. She confronts the reader with a great variety of activated scripts and reduces the action structures to a few movements, so that the application of familiar action sequences becomes very difficult. This forces readers—even more so than in the case of Gonne—to call their own scripts into question and to make adjustments.
Herman avoids exaggerated statements on the differences between the genres to which his examples belong, and quite rightly so. A novel does not need to be more demanding than an autobiography, and authors of children’s literature may be less than serious with respect to their didactic assignment. Nevertheless, this synchronic approach provides more insight into the position of a narrative text within the genre system. The combination of synchronic and diachronic approaches enables postclassical narratology to contribute to literary history.
Which scripts do Mutsaers and Krol activate and how do they use them? “Pegasian” immediately invokes our expectations about a conversation. Conversations can develop in many directions, but in this case the possibilities are limited by the fact that one of the participants seems to be in a position of power. All of us have at one time been addressed by an insistent and authoritarian figure, and thus we are familiar with this kind of conversation. Our knowledge about its ending will partly depend on our own experiences—those who have suffered because of powerful people may find it hard to imagine a good result. This could mean that important aspects of the text are neglected to the benefit of personal projection. Nevertheless, a lot of narrative prose possesses a power of activation that runs at least partly counter to this projection. This is also the case in “Pegasian.” The girl’s first utterance already indicates that she will put up some resistance, and so we immediately have to integrate this element into our expectations. How do conversations between an authoritarian figure and an assertive younger person develop? It is unlikely that the person in power will simply back down, but it is clearly possible that he (or she) might have recourse to instruments other than verbal ones. Younger persons may give in after a while, but their resistance may also continue, especially if an authoritarian figure’s irritation shows through and if younger persons facing such figures notice that, in spite of their subordinate position, they have achieved some results.
The first part of “Pegasian” largely conforms to the conventional development of an argument between an authority figure and a pupil. The text’s first sentences call for such a script. The riding master slowly loses patience with the girl but eventually vents the accumulated frustration on the horses instead of the girl: “These horses are moving around like turtles. Time to bring out the whip.” The pupil, on the contrary, enjoys resisting the riding master and stands her ground. With respect to script theory, it is interesting that the text strongly downplays the argument script in the second part. In order to understand the girl’s eventual insight, the reader can turn to scripts on adolescent behavior, some more condescending than others. Perhaps the riding master has convinced the girl after all, and perhaps her resistance was simply due to the idea that this is the way one reacts when one is young. Or perhaps her eventual insight suggests that it would have been better not to put up any resistance at all instead of making a scene like a typical teenager. These interpretations obviously depend on the reader’s conception of an adolescent.
The argument part of the script is activated by the conversation, the reconciliatory part by the story’s ending. Of course the text’s formal aspects influence the application of the activated scripts. Multiple focalization and a reclusive narrator make it hard to interpret the word “finally” in the second part: “Finally, she understands.” If “finally” expresses the narrator’s evaluation, the latter seems to share the disdain that is part of the condescending script about willful teenage behavior. If the word must be assigned to the girl, this triggers the slightly disarming suggestion that she was willing to understand the usefulness of the riding breeches but that she simply was not convinced.
The first paragraph of “The Map” describes Mr. and Mrs. Paalman’s bookstore in the village of Dorkwerd. A spatial description may trigger expectations as to the events to come in the environment at hand. In Krol’s story the bookstore, as an example of the blinding effects of Christianity, can be used as the starting point for a number of different scripts. Perhaps the main character will try to steal books in the shop. Perhaps Mr. and Mrs. Paalman will have an adventure inspired by their contact with books. Perhaps the world of books will liberate the couple’s imagination, which might lead to a clash with their environment. On the basis of the sentence “he’s nice, she’s wearing the pants,” the reader may suspect that the bookstore will be the scene of a battle between the sexes.
The fact that so many scripts are imaginable indicates that the text has a high degree of narrativity. Aspects of these scripts can be integrated with the scripts evoked by the second paragraph. At the beginning of this paragraph, the first-person narrator introduces himself and emphasizes the importance of the shop window being shaded. He sees the map when he is not supposed to see it, which gives the object an extraordinary value. Will there be a confrontation between the first-person narrator and the ideology symbolized by the shutters? Will the map show him the way out of the community dominated by Christian prohibitions? The adventure suggested by these scripts is limited, however, to trips to places that are on the map but that had not been visited by the main character before. Liberation will not come as fast as might have been expected, because the experiencing I keeps returning home during his map period. The end of his narration reinforces this qualification since the first-person narrator tells us the map had become superfluous and that he therefore did not keep it.
The liberation script, which was evoked by, among other things, the windows being shaded on Sunday, vanishes along with the map. The undermining of this script leads the reader to reinterpret the first-person narrator’s entire development along the lines of a much less adventurous script. This disappointing development is also a script that appears to be inherent in growing up. The child cherishes dreams and imagines scripts that can in fact be realized only to a very limited degree. Krol’s choice of a first-person narrator intensifies the reader’s identification with the boy, which frustrates the reader’s expectations even more. This disillusionment can contribute to the insight that the reconstruction of the past is distorted by relativization and disappointment.
Ralf Schneider has sought to theorize “the reader’s construction of a mental model in the process of understanding character.”289 In narrative processing, readers integrate information about characters from the text (bottom-up) and information from memory storages, including frames and scripts (top-down). The result is a mental model that is “continually updated, modified or revised to adapt to ongoing information processing.”290 Schneider submits that there are essentially two kinds of models the reader can construct, depending on the quality and the quantity of information from the two sources. Readers may either land on a character “category” (which happens most often top-down, though narrative texts themselves can of course make explicit suggestions in this regard), or they resort to “personalization” (which happens most often bottom-up, although this might be enhanced by the top-down expectation that literary texts present to individuals—an expectation that defies categorization).