A New Anatomy of Storyworlds: What Is, What If, As If 9780814215081, 0814215084, 9780814282274, 081428227X, 9780814258415, 0814258417

Citation preview

A N E W A N AT O M Y O F S T O R Y W O R L D S

T H E O R Y A N D I N T E R P R E TAT I O N O F N A R R AT I V E

James Phelan, Katra Byram, and Faye Halpern, Series Editors

A NEW ANATOMY OF STORY WORLDS WHAT IS, WHAT IF, AS IF

Marie-Laure Ryan

T H E O H I O S TAT E U N I V E R S I T Y P R E S S C O LUM BU S

Copyright © 2022 by Marie-Laure Ryan. All rights reserved. Published by The Ohio State University Press. Library of Congress Cataloging-in-Publication Data Names: Ryan, Marie-Laure, 1946– author. Title: A new anatomy of storyworlds : what is, what if, as if / Marie-Laure Ryan. Other titles: Theory and interpretation of narrative series. Description: Columbus : The Ohio State University Press, [2022] | Series: Theory and interpretation of narrative | Includes bibliographical references and index. | Summary: “Uses a broad set of works—from Sokal’s hoax, to Maus, from Saussure to Barthes, from Kafka to virtual reality—to ground narratology in the concept of world and interrogate key subjects such as narrator, plot, character, fiction, mimesis, and diegesis”—Provided by publisher. Identifiers: LCCN 2022018894 | ISBN 9780814215081 (cloth) | ISBN 0814215084 (cloth) | ISBN 9780814282274 (ebook) | ISBN 081428227X (ebook) Subjects: LCSH: Narration (Rhetoric) | Fiction—History and criticism. | Discourse analysis, Narrative. Classification: LCC PN212 .R933 2022 | DDC 808.3/9—dc23/eng/20220524 LC record available at https://lccn.loc.gov/2022018894 Other identifiers: ISBN 9780814258415 (paper) | ISBN 0814258417 (paper) Cover design by Nathan Putens Text composition by Stuart Rodriguez Type set in Minion Pro

CONTENTS

List of Illustrations

vii

Acknowledgments

ix

INTRODUC TION

Grounding Narratology in the Concept of World

1

CHAPTER 1

Truth: Discourse Types and Theories of Truth

19

CHAPTER 2

Fiction: The Possible Worlds Approach to Fiction and Its Rival Theories

34

CHAPTER 3

Narrator: Decomposing a Theoretical Primitive

58

CHAPTER 4

Characters: Textual, Philosophical, and “World” Approaches to Character Ontology

74

CHAPTER 5

Plot: Cheap Plot Tricks, Plot Holes, and Narrative Design

90

CHAPTER 6

Mimesis and Diegesis: Complementing Each Other

110

CHAPTER 7

Parallel Worlds: Physics, Narrative, and the Multiverse

123

CHAPTER 8

Impossible Worlds: Dealing with Logical Contradiction

145

CHAPTER 9

Virtual Worlds: Narrative and VR Technology

162

CHAPTER 10

Transmedia Worlds: Industry Buzzword or New Narrative Experience?

182

Bibliography

205

Index

219

I L LU S T R AT I O N S

FIGURES FIGURE 1

Gradual and binary theories of fiction

53

FIGURE 2

How narrator theories group together fictional forms

62

FIGURE 3

Degrees of characterhood

88

FIGURE 4

Mimetic and diegetic narratives

114

FIGURE 5

Mimetic and diegetic narration in Maus, by Art Spiegelman

119

FIGURE 6

Three types of counterpart relations in many-worlds narratives 138

TABLES TABLE 1

The distribution of narratorial functions in actual agents

69

TABLE 2

The distribution of narratorial functions in fictional agents

71

TABLE 3

Types of immersion in various media

vii

170

ACKNOWLEDGMENTS

I dedicate this book to Thomas Pavel, whose interest in my work when I was a young scholar was a tremendous encouragement and whose own work had a lasting influence on mine, and to the late Emma Kafalenos, a spirited and refreshingly opinionated friend, whose enthusiasm for all things narratological and support of my work was a major inspiration for an independent scholar like me to persevere with this strange hobby of narrative theory. Some of the chapters of this book originated in ideas that came spontaneously to me, or rather whose origin I cannot remember. Others are the latest outgrowths of long-standing interests of mine. Still others originated in topics that were imposed on me by invitations to participate in scholarly communities; these chapters are no less important to me than the spontaneous and longtime-interest ones, because they forced me to learn something new rather than rely on my pre-existing “narratological capital.” For stimulating my mind through their discussion, I thank: Carlos Reis of Coimbra University (for chapter 4), Science, Text, Audience, Reception (STAR) (for chapter 7), the 2018 Lisbon Consortium Summer School (for chapter 9), and Karl Renner of the Gutenberg University in Mainz, Germany, who invited me to spend a year there as a Gutenberg Fellow (for chapter 10).

ix

x  •  Acknowledgments

My thanks also go to the editors of the Theory and Interpretation of Narrative series and to the anonymous reader of the manuscript. Their careful reading and insightful suggestions have greatly improved the readability of this book. The chapters are based on the following publications. All of them have been extensively revised and updated to include more recent research. Chapter 1: “Truth Without Scare Quotes: Post-Sokalian Genre Theory.” This article first appeared in New Literary History 29, no. 4 (Autumn 1998): 811–30. Copyright © 1998, The University of Virginia. Published with permission by Johns Hopkins University Press. Chapter 3: The second half is based on “Narrator: Decomposing a Theoretical Primitive,” Narrative 9, no. 2 (2001): 146–52. Published with permission of The Ohio State University Press. Chapter 4: “Textual, Philosophical and ‘World’ Approaches to Character Ontology,” Neohelicon 45 (2018): 415–29. Published with permission of Springer Verlag. Chapter 5: “Cheap Plot Tricks,” Narrative 17, no. 1 (2009): 56–75. Published with permission of The Ohio State University Press. Chapter 7: “From Parallel Universes to Possible Worlds: Ontological Pluralism in Physics, Narratology and Narrative,” Poetics Today 24, no. 7 (2006): 633–74. Published by permission of Duke University Press. https://www.dukeupress.edu. Chapter 8: “Impossible Worlds,” in The Routledge Companion to Experimental Literature, edited by Joe Bray, Alison Gibbon, and Brian McHale, 368–79, 2012. Copyright © 2012. Reproduced by permission of Taylor & Francis Group. Chapter 9: “Narrative in Virtual Reality: Anatomy of a Dream Reborn,” Facta Ficta: Journal of Theory, Narrative and Media 2, no. 2 (2018). Chapter 10 combines materials from “Transmedia Storytelling and Transfictionality,” Poetics Today 34, no. 3 (2013): 362–88, republished by permission of Duke University Press, https://www.dukeupress.edu, and “Transmedia Storytelling and Its Discourses,” published in Transmediations: Communication Across Media Borders, edited by Lars Ellestrøm and Niklas Samose, 17–30, London: Routledge, 2020. Copyright © 2020. Reproduced with permission of Taylor & Francis Group.

INTRODUCTION

Grounding Narratology in the Concept of World

This book explores some of the implications of the concept of world for narratology through what I call the what is, what if, as if approach. It may seem intuitive that narrative texts, whatever their medium, offer a world to the user’s imagination, a world inhabited by individuals (i.e., characters) who bring changes to its global state through their actions or who react to changes due to accidental events. But as self-explanatory as the concept of world may appear, it has been subjected to multiple interpretations, and it has traveled across many disciplines: from cosmology to philosophy to digital culture to popular culture to literary theory. Before I address its potential for narratology, I propose a quick tour of its manifestations in these other domains.

The Traveling Concept of World Though the cosmological sense of world can be regarded as literal, this sense is not without ambiguity. The Oxford English Dictionary (OED) defines world as: 1. The earth. 2. Another planet like the earth. 3. The material universe of all that exists; everything. This definition leads—in the endless deferral of meaning typical of dictionaries—to universe: “All existing matter and space considered as a whole; the cosmos.” If the universe is “everything that exists,” it is made of many worlds in sense 2 above, but there is only one universe. Yet the 1

2  •  I ntroduction

definition of universe is followed in the OED by some facts that describe it as a particular place or entity rather than as a totality encompassing everything: it is ten million light-years in diameter; it contains a vast number of galaxies; it was created by the big bang thirteen billion years ago; it is steadily expanding. This description suggests that there could be other universes with different properties: they could be bigger or smaller, contain a different number of galaxies, and be born from a big bang that took place at a different time, or from a phenomenon other than a big bang. Since these other universes would exist materially, the totality that includes everything would be not the universe but a multiverse, a notion that has recently gained considerable traction in the scientific community (Rubenstein 2014). The number of worlds in sense 2 in the multiverse is absolutely mind-boggling. While astronomers discovered more and more worlds in senses 2 and then 3, the idea of plurality of worlds migrated from cosmology to philosophy. Leibniz argued in the Theodicy that God considered all possible worlds, and chose the best one to be actualized, even though this “best possible world” still contains a lot of suffering. But while the mind can contemplate an infinity of worlds, Leibniz’s cosmology remains a one-world model since only one of them has been called into existence by God. Leibniz’s distinction between the actualized and the merely possible foreshadows the central tenet of possible worlds (henceforth PW) theory, despite the claims of its proponents that the theory is not a metaphysical model but more modestly a tool in the service of logic. PW theory was originally developed by members of the school known as analytic philosophy, such as Saul Kripke, David Lewis, Nicholas Rescher, Alvin Plantinga, and Jaakko Hintikka, as a way to solve problems in the attribution of truth-values to propositions modified by so-called modal operators (it is possible/impossible/necessary that p; it is good/bad/indifferent that p; p is allowed/prohibited/mandatory) and to counterfactuals. While the theory postulates an infinity of worlds, corresponding to ways things could be or could have been, it does not grant equal status to all worlds: one of them is the actual world (AW), while all others are nonactual possible worlds (NAPWs). The truth-value of propositions can be assigned separately for each world of the system. A proposition is necessary when it is true in all worlds; possible when it is true in some worlds but not others; and impossible when it is false in all worlds. Moreover, a proposition that is false in the actual world may be true in a NAPW (for instance, that Darth Vader exists), and, vice versa, something that is true in the AW (that Donald Trump exists) may be false in a NAPW, such as the world of Star Wars. The opposition of an actual or real world to merely possible ones is central to PW theory, but the exact nature of actuality is open to debate. Two theories

G rounding N arratology in the Concept of W orld  •  3

prevail. According to one of them, the actual world is the only one that exists independently of the human mind; all others are constructs of the imagination, such as dreams, hallucinations, and fictional stories. In the other interpretation, known as modal realism and defended by David Lewis, all worlds exist absolutely, and the difference between AW and NAPW is merely a matter of point of view: actual is an indexical term, like I or here or now, whose reference depends on the speaker (1979, 184; see also chapter 2 in this volume). For me the actual world is planet Earth and the cosmic system of which it is a part, but for Darth Vader the actual world is the world of Star Wars, and he rightly regards himself as a real person, not as a fictional character. More recently the concept of world has become associated with the creations of the imagination that media bring into our lives, and with the technologies that support these media. As media-produced representations, worlds should no longer be defined as “everything that exists” but should rather be conceived in cognitive or phenomenological terms, as corresponding to the extent of an individual’s experience. “My world” in this sense is a habitat, an environment (German: Umwelt) made of the objects and of the people that surround me and that define my mental horizon. It is a small, subjective totality located within the large, impersonal totality of everything that exists. This conception of world finds a moving visual expression in Andrew Wyeth’s painting Christina’s World, which depicts a semiparalyzed woman crawling through a field with farm buildings in the distance, painfully striving toward a horizon she will never reach. The title suggests that this landscape represents the stage setting of her life drama, everything that she will ever know. The virtual worlds of video games have truly become worlds in this phenomenological sense. They not only engage the muscular and strategic abilities of players, as do sports games and board games, they can also stimulate the narrative imagination through the tasks and scripted quests that are offered to players. In online multiplayer gameworlds such as Second Life or World of Warcraft, players identify with avatars, form communities, build objects, and, through all these activities, co-create virtual worlds. Insofar as the world-creating ability of media leads to a sense of being there in the displayed environments, the concept of world becomes tightly connected to that of immersion, an experience as central to narrative media such as film and novels as it is to computer games. Even when the medium does not allow physical interaction, as is the case with novels and films, people may participate spontaneously in the creation of imaginary worlds through external contributions such as fan fiction, art, and videos. Any immersive world can be made interactive by dedicated fans. While the vast majority of fictions can be said to build a world, a distinctive cultural tradition has focused on world-building as an autotelic project

4  •  I ntroduction

that puts into practice J. R. R. Tolkien’s notion of secondary creation as imitation of God’s primary world-creation. Mark J. P. Wolf (2012) has identified over 1,440 imaginary worlds in literature and other media, a list from which he later extracted twenty-five canonical ones (appendix to Wolf 2017, 348–62). The most notorious examples include Kallipolis in Plato’s Republic (circa 380 BCE), the islands and celestial bodies of Lucian of Samosata’s True History (first century), the islands of The Book of Sir John Mandeville (circa 1357), Utopia from Thomas Moore (1516), the lands of Jonathan Swift’s Gulliver’s Travels (1726), Lewis Carroll’s Wonderland (1865), Edwin Abbott’s Flatland (1884), L. Frank Baum’s OZ (1900), J. R. R. Tolkien’s Middle Earth (from 1937), George Lucas’s Star Wars universe (from 1977), and the online worlds created by the users of Linden Lab’s Second Life (from 2003). Spurred by the indefatigable work of Mark J. P. Wolf, a thriving field of investigation focuses on these imaginary planets, celebrating the inventiveness that goes into their creation and bridging the gap between popular and literary culture (Wolf 2012, 2017, 2018).

Worlds in Literary Theory and Narratology In standard literary criticism, the notion of world has long been used in an informal way to describe the distinctive character of an author’s creative activity, the inner vision that manifests itself either in an individual text or in the author’s total production. There is a world of Proust, a world of Homer, or a world of Jane Austen. But as long as literary theory remained focused on the signifier, at the expense of the signified, this informal use of world could not gel into a theoretical concept. From New Criticism to deconstruction, the “textualist” critical movements inspired by the linguistic turn of the fifties and sixties were too focused on language, on écriture, to pay attention to the concept of world, but if they had done so, they would have conceived world as an infinite sum of meanings that could not be paraphrased (a favorite battle 1 cry of New Criticism was indeed “The heresy of paraphrase” ), and that could not be separated from the text. It follows from these positions that the text is the only mode of access to its world. And because textualism is reluctant to isolate a narrative level of meaning from the global textual world, it implicitly adhered to a strict formula: one text—one world—one story. After the linguistic turn came the narrative turn of the eighties, and narrative or story became prominent. One of the effects of the narrative turn was a shift of attention that, without ignoring the signifier, restored the importance

1. Title of chapter 11 of Cleanth Brooks’s The Well Wrought Urn.

G rounding N arratology in the Concept of W orld  •  5

of the signified. While stories are transmitted by discourse, which means by text, they remain inscribed in our minds long after the signifiers have vanished from memory. This means that story is a cognitive rather than a linguistic construct. The fact that stories can be summarized, adapted, and translated, and that they can be told by various media, emancipates them from language and makes them somewhat independent from the particular signs through which they are transmitted. The structuralist idea that Cinderella and a Chinese folk tale could be versions of the same story, which was heretical for textualism (Smith 1981), becomes very acceptable for a narratologist. Instead of one text, one world, one story, one could now have the possibility of many texts—one world—one story, or, if one assumes that every different version of a story creates a distinct world, many texts, one story, many worlds. The next step in this development is the emancipation of the concept of world from the concept of story. If we conceive world as a spatiotemporal environment made of a sum of existents that the story modifies through its constitutive events, then there is no reason to limit worlds to one story, because several distinct sequence of events can take place simultaneously or successively in the same environment. The phenomenon of transfictionality (SaintGelais 2011) demonstrates that existing storyworlds can be expanded through new stories. While transfictionality is arguably as old as fictional story­telling itself, contemporary culture, whether popular or highbrow, implements the full range of possible relationships between text, world, and story. This play with worlds can take the following forms: • Narrative proliferation (many texts, one world, many stories): for instance, the many novels of Balzac’s Comédie Humaine or the many Star Wars stories take place in the same world. • Ontological proliferation (one text, many worlds, many stories, or a story that spans multiple worlds): this case is represented by what David Bordwell (2002) calls “forking path narratives”: a character is facing a decision between different courses of action, leading to different worlds, and each of these courses of action is narrated. Example: the film Run Lola Run. Another way for a text to contain many worlds is to represent a metaleptic communication between distinct ontological domains, such as a world presented as real and a world presented as fictional (Woody 2 Allen’s film The Purple Rose of Cairo).

2. By contrast, a text that represents space travel may include many planets, but only one world, if by world one understands “all that exists.”

6  •  I ntroduction

• Medial proliferation: many texts, one world, one story retold through different media, or, in a transmedial version of narrative proliferation, many different texts of different media telling different stories about the same world. This is the phenomenon currently known as transmedia storytelling (Jenkins 2006). • Transworld migration: a character from a known fiction being placed in a different world, where he or she may encounter characters from other pre-existing worlds (the comic book series The League of Extraordinary 3 Gentlemen, by Alan Moore and Kevin O’Neill ).

Narratology deals with such phenomena through the concept of storyworld. Though storyworld has been used earlier by other critics, its current prominence is largely due to David Herman’s attempt to define the concept and to situate it within a cognitive framework. While storyworlds are, broadly speaking, the worlds evoked by stories, narrative texts, conversely, are “blueprints for a specific mode of world-creation. Mapping words (or other kinds of semiotic cues) unto worlds is a fundamental—perhaps the fundamental— requirement for narrative sense-making” (2009, 105). Herman admits, however, that “this mapping operation may seem so natural and normal that no ‘theory,’ no specialized nomenclature or framework of concepts is necessary to describe and explain the specific procedures involved” (105). To construct a storyworld is simply to form a mental representation of a narrative text; users only become aware of the operations involved in this construction when these operations are blocked, as happens in postmodernist antinarrative texts. Yet even though the self-evidence of storyworlds seems to make definition dispensable, Herman does not give up on the attempt to capture the nature of storyworlds: they are “global mental representations enabling interpreters to frame inferences about the situations, characters, and occurrences either explicitly mentioned in or implied by a narrative text or discourse” (106). As the target of the interpretive activity elicited by narratives, storyworlds transcend the distinction between fictional and factual narratives: even when they purport to represent the real world, narratives produce a storyworld, which may or may not correspond to the reference world. Moreover, if the ideal narrative experience is to be described in terms of immersion (Schaeffer 1999, Ryan 2001), storyworld functions as the milieu in which immersion takes 3. The characters in The League of Extraordinary Gentlemen include Mina Harker, from Bram Stoker’s Dracula; Campion Bond, grandfather of Ian Fleming’s James Bond; Allen Quatermain, from H. Rider Haggard’s King Solomon’s Mines; Hawley Griffin, from H. G. Wells’s The Invisible Man; Dr. Jekyll and Mr. Hyde, from Robert Louis Stevenson’s eponymous novel; and Prof. Moriarty, from Arthur Conan Doyle’s Sherlock Holmes stories.

G rounding N arratology in the Concept of W orld  •  7

place. To be immersed in a narrative means to transport oneself in imagination to its storyworld, whether or not this world is regarded as an image of the real world. And yet, for all its apparent self-evidence, the concept of storyworld is not universally accepted. We will see in chapter 2 conceptions of fiction that do not rely on the idea of fictional world. Richard Walsh argues that fictionality is a rhetorically distinct mode of communication (distinct from commitment to factuality) and that regarding narrative fiction as world-creation is incompatible with communication: “It is not just that fictional-worlds approaches have nothing to say about communicative purposes [which he conceives as the expression of messages with real-world relevance]; it is that they actually foreclose the possibility that the distinctiveness of fiction might have something to do with its communicative use” (2019a, 401). Moreover, Walsh believes that the essence of narrative lies in temporality, and that worldmaking is not only “not intrinsic to narrative, but in fact fundamentally anti-narrative,” arguably because he regards worlds as spatial constructs (2019b, 521–22). And, finally, he rejects the characterization of the (ideal) narrative experience as immersion. I refer the reader to Walsh (2019a, 2019b) to find out his reasons for such pronouncements (for my response, see Ryan 2019a, 2019b). My point here is not to get to the root of these attempts to eliminate worlds from narratology and/or theories of fictionality but to suggest that the concept of storyworld must be defended rather than taken for granted.

The What Is, What If, As If Approach Through the formula what is, what if, as if, I hope to inspire reflection on the ontological and experiential (or phenomenological) implications of story­ worlds. Narrative texts cannot represent “all that exists,” but they can, and should, represent how individual existents relate to the people and objects that define their living environment. This means that for a text to be narrative, it must give a partial, but solid answer to the question “what is?” by evoking to the imagination a collection of entities—characters, objects, events—that exist in some world, that are causally interconnected, and that either undergo or cause transformations. As I write elsewhere, “Worlds can be thought of in two ways: as containers for entities that possess a physical mode of existence (events can be considered such entities because they affect solid objects and are anchored in time and space) and as networks of relations between these entities” (Ryan 2019b, 63). Not all texts are narratives, because not all texts project worlds inhabited by individuated entities. Philosophical texts deal

8  •  I ntroduction

with ideas, not with concrete existents; as for scientific texts, they do concern concrete natural objects, but most of them are not narratives, or are only weakly narrative, because their purpose is to make general statements about the behavior of entire categories, such as all atoms or all monkeys. Among storyworlds, an ontological distinction must be made between what-is and what-if worlds. What is refers, obviously, to factual texts that purport to represent the real world, while what if describes fictional texts that create imaginary worlds located at variable distances from the real world. But the ontological distinction between what is and what if also operates within both kinds of worlds. Insofar as fictional texts create their own worlds, they bring objects that do not exist in reality into imaginative existence. Just as horses exist in the real world but unicorns do not, Emma Bovary exists fictionally within the world of Flaubert’s novel but the characters of the novels that nourish her imagination do not. Sorting out “what is” from “what is not” is a fundamental interpretive operation for all storyworlds, whether fictional or factual. On the other hand, if we interpret “what if ” as the ability to take into consideration counterfactual scenarios and possible hypothetical developments out of a given situation, it is a key aspect of mental life that presents immeasurable adaptive advantage. On this account, thinking involves the profoundly narrative process of taking us beyond what is. As Brian Boyd puts it, “A mental architecture that processes only true information remains severely constricted. Most discovery involves supposition. We cannot even think seriously about cause without counterfactuals: if this had not happened, would that still have occurred” (2009, 198). While what-if thinking is essential to practical life, narrative fiction can be regarded as an extension of this ability that grants autonomy and therefore aesthetic value to the creation of imaginary worlds. No longer subordinated to real-world action, the what-if worlds of fiction become objects of enjoyment worth contemplating for their own sake. Within fictional worlds, what-if thinking operates in exactly the same way that it operates within the real world. For instance, as the philosopher Georg Henrik von Wright has shown, the decision to take action, and the interpretation of other people’s actions, requires the mental contemplation of two alternative sequences of events: what the agent intends to accomplish, and what the agent thinks would happen without taking action. This reasoning applies to literary characters as well as to real people because assuming that characters have the same reasons for acting as we do (Schaeffer and Vultur 2005)4 is a fundamental law of narrative sense-making. Understanding narratives, whether fictional or factual, thus requires the mental simulation of 4. Paraphrasing Ricoeur, Schaeffer and Vultur define basic mimeticism as a logic of narrative emplotment that “mirror[s] the implicit inferential logic of action guiding our real-world experiences” (2005, 310).

G rounding N arratology in the Concept of W orld  •  9

the virtual sequences of events, or possible worlds, contemplated by the characters. As I have suggested in Ryan 1991, the diversification of these virtual sequences, which I called virtual embedded narratives, plays a crucial role in making a story tellable. To take a familiar example: the pathos (and tellability) of Oedipus Rex resides not in the fulfillment of the prophecy that he will kill his father and marry his mother but in the fact that this fulfillment is brought about by the very attempt to prevent it. There are consequently two competing sequences of events, one realized—the prophecy—and the other virtual, its prevention. While the what-if worlds of fiction are ontologically different from the what-is worlds of factual narratives, they become what-is worlds to the imagination through the operation of an as-if game of pretense. I have described this game as an imaginative recentering into fictional worlds, an operation through which they become real in make-believe (Ryan 1991). In its minimal form, the as-if effect is the acceptance of the narrative as true of its reference world (with needed adjustments for narratorial unreliability), while in its most extreme form, it becomes an immersive sense of presence of the storyworld. By insisting on the pseudoreality of fictional worlds, the as-if principle exemplifies the realistic bias that advocates of unnatural narratology regard as a shortcoming of the discipline. Realism in this case must be taken not as a style of representation, nor as a focus on certain narrative themes, nor even as a close resemblance between the real world and the storyworld, but in the truly ontological sense of asserting the existence of a world containing characters, states, and events, and asking users to imagine this existence as independent of the medium that represents it, even though they know that in the case of fiction it is a product of this medium (and of the mind that designs it). I believe indeed that narrative is a fundamentally realistic mode of representation, and that the self-reflexive, illusion-destroying strategies that form the focus of unnatural narratology are parasitic upon this realism. Before postmodernism and its antinarrative games of takeaway came along, there was aesthetic illusion.

Situating My Approach within Contemporary Narratology Through this emphasis on worlds, I wish to distinguish my approach from four dominant schools of narratology that are featured in Narrative Theory (henceforth NT), a highly stimulating and original book published in 2012 by The Ohio State University Press: the rhetorical approach, advocated by James Phelan and Peter Rabinowitz; the feminist approach, whose champion is Robyn Warhol; an approach by David Herman that he defines as a focus

10  •  I ntroduction

on “the nexus of narrative and mind” but which most readers will associate with cognitive narratology; and the “unnatural” approach, pioneered and represented by Brian Richardson. Though the authors recognize in the preface that “the four approaches discussed here do not cover the entire field and that there are many theorists doing valuable work that does not fit comfortably under any one of the four rubrics we use to describe our work” (xi), these approaches have come to dominate the narratological scene, undoubtedly thanks to the importance of the work of their respective advocates, but also because of the publicity they receive by being featured in a book titled Narrative Theory, and because of the convenient labels through which they are identified: never underestimate the power of brand names. Through the collection of essays presented here, I hope to validate the authors’ claim that there can be valuable work outside of their approaches. By featuring these four different schools, NT puts into practice the expansion of the so-called classical phase of narratology, associated with structuralism (Barthes, Genette, etc.) into what David Herman, at the end of previous century (Narratologies, 1999), called a postclassical phase, a term that has become viral in the small world of narratology. Postclassical narratology covers developments such as these: 1. Extending the corpus of narratology to cover nonliterary narratives (factual, conversational). 2. Extending the corpus to nonverbal media (film, drama, graphic novels, digital narratives). 3. Dividing narratology into distinct “narratologies” (feminist, postcolonial, queer, environmental, unnatural, etc.). 4. Revisiting classical notions (author, narrator, focalization, description, action, causality, the mimetic/diegetic distinction). 5. Paying more attention to pragmatic issues, such as context and purpose of transmission. 6. Developing interest in fiction/fictionality (an issue previously considered philosophical territory). 7. Blurring the distinction between theory and interpretation.

I endorse most of these developments, but I remain skeptical toward 3 and 7. This reservation explains why I wish to keep my distance from one of the four schools, namely feminist narratology. While there are many feminists who use narratological concepts, often brilliantly, I believe that there is no such thing as a feminist narratology, if by this term one understands a separate narratology needed to address questions

G rounding N arratology in the Concept of W orld  •  11

of gender in narrative texts. There is instead a narratology (or narratologies, to account for theoretical disagreements and alternative models) that can be used in support of feminist (as well as queer, postcolonialist, ecocritical, etc.) readings of narrative texts, to detect tendencies in texts written by women, or to describe narrative genres favored by women, such as soap operas, romance, or so-called chick-lit. On the issue of feminist narratology, I stand with Nilli Diengott of the Tel Aviv school of poetics, who argued in 1988 that the idea of a feminist narratology would transgress the boundary between interpretation and poetics. She claimed that narratology should attempt “to get away from categories that are clearly ‘semantically loaded’” (48) and that “determining meaning in narrative is what interpretation, not theoretical poetics, is all about” (45). The alternative to regarding feminist narratology as a formal system in its own right and as an alternative to classical narratology is maintaining a distinction between genuine narratological concepts, that is, concepts designed for the study of narrative texts and the concepts proposed by feminist or gender theory, such as patriarchy, gender as performance, nonbinary conceptions of gender, or intersectionality, whose relevance is broader than the study of narrative. The most famous proposal for a feminist revision of classical narratology has been Susan S. Lanser’s suggestion that the gender of the narrator is “a necessary and important element of a formal—i.e. structural and descriptive—poetics of narrative” (1999, 172; italics original), and that, in the case of impersonal narration, the gender of the author is the gender of the narrator (2015, 30), a principle known as the Lanser rule. While some readers may occasionally imagine anonymous third-person narrators as gendered (it is certainly not my intent to legislate over the reader’s act of imagination), failing to do so is certainly not a case of misreading, and Lanser’s rule stands in blatant contradiction to the narrator-optional (Patron 2016, 2020) or disembodied-narrator theories discussed in chapter 3. To summarize my position, I am not proposing to expel the term feminist narratology from the terminology of the discipline but instead proposing to conceive the task of feminists working with narrative texts along the lines suggested by Robyn Warhol: “Feminist narratologists have not proposed comprehensive models of narrative that would contradict or supplant those which have developed in the larger field of narratology . . . Feminist narratologists tend most often to combine the insights of feminism and of narratological analysis in developing gender-­centered interpretations of individual texts” (2005, 162). Rhetorical narratology, in contrast to the feminist kind, has developed a number of original concepts that are widely used in narratology. Wayne Booth, the founder of the movement, proposed the largely undisputed unreli-

12  •  I ntroduction

able narrator (though opinions differ as to what exactly should be regarded as unreliable narration) and the more controversial implied author, to which I return in the next paragraph. Phelan (1989) added a distinction between three dimensions of characters (synthetic, mimetic, and thematic), and in NT, Phelan and Rabinowitz contribute further distinctions between three functions of narrators: report, evaluate, explain (34), and three readerly responses: interpretive, ethical, and aesthetic (6). The rhetorical approach’s forte lies in the analysis of the layered nature of fictional communication, as well as in the description of the dynamics of plot. But I have two theoretical reservations to the approach. The first has to do with the authors’ definition of narrative as communication: “Narrative is somebody telling somebody else, on some occasion, and for some purposes, that something happened to someone or something” (NT 3; italics original). This characterization works well for factual narratives, but in the case of fictional narratives, should it apply to the actual level of author-audience communication, or to the imaginary level of the narratorial utterance? Communication between authors and readers occurs by displaying a narrative, rather than by telling it, if by telling one understands informing the reader that something happened. On the narratorial level, occasion and purpose are rarely definable: we cannot identify what motivates the narrator of Mrs. Dalloway to tell the story. Even in some instances of first-person narration, for instance the discourse of Benjy in William Faulkner’s The Sound and the Fury, the narration cannot really be considered an act of communication fulfilling a specific purpose and performed on a specific occasion. My other difference with the rhetorical approach lies in my skepticism toward the notion of the implied author, which Phelan and Rabinowitz vigorously endorse (NT, for instance, p. 31). I have stated elsewhere (Ryan 2011)5 my reasons for rejecting the notion. Here they are in a nutshell: I believe that the implied author is simply the author as imagined by the reader on the basis of the text and possibly additional information, and that the qualifier implied can be deleted from all uses without significant interpretive implications. If there is a need for distinguishing an implied from an actual source, it occurs in factual rather than in fictional communication because this implied source (speaker, writer, author) is the entity that fulfills what Searle (1969) has called the utterance’s felicity conditions. The distinction between actual and implied speaker is therefore necessary to detect lies, which do not occur in fiction since the text is a performative act of world-creation.6 5. Available on my website: http://marilaur.info/impauthor.pdf. 6. Felicity conditions are also broken in the case of error, but since the speaker believes the statement to be true, the implied speaker does not differ from the actual speaker.

G rounding N arratology in the Concept of W orld  •  13

Though the proponents of unnatural narratology (Alber, Iversen, and Nielsen in addition to Richardson; see Alber et al. 2012) hold different views of what is unnatural about some narratives, they agree that the contrast between natural and unnatural is a matter of mimeticism. Not all narratives are mimetic, but narratology is supposedly unable to deal with the nonmimetic kind because of a realistic or mimetic bias; it will therefore take an expansion, or revision of narratology to account for the full spectrum of narratives. Since the alleged bias of narratology lies in its reliance on the notion of mimesis, or representation, we need to take a closer look at this concept.7 The scope of mimesis can be understood in different ways: 1. The representation of something that happened in the actual world. By this definition, mimeticism is the defining property of truthful nonfiction. 2. The representation of what could (but did not) happen in the real world. This interpretation associates mimesis with realistic fiction. 3. The representation of what did happen in some possible world that differs to a variable extent from the real world. This interpretation extends mimeticism to the fantastic and to science fiction, but it excludes impossible worlds.

To this list I would like to add a conception inspired by Schaeffer and Vultur (2005, 309–10), who themselves rely on Paul Ricoeur: 4. The ability of a text to inspire the imagination to create mental images of concrete objects and processes, such as characters, settings, and events, and to ascribe to characters roughly the same reasons for acting as we do.

Richardson’s concept of the unnatural excludes 1 (as the prototype of the natural) and 2 (as a fictional imitation of 1). Within the unnatural, he distinguishes two categories: the nonmimetic, which corresponds to category 3, and the antimimetic, which is represented by experimental texts, especially postmodern ones. The radically antimimetic texts that Richardson is so eager to recuperate for narratology are not only antimimetic but, as Herman writes in his comments on Richardson, “antinarrative—that is, purposefully designed to thwart the worldmaking process” (NT 223). Narrativity is not a binary, either/ 7. Here I take mimesis in an Aristotelian way, as meaning representation. In chapter 6, I concentrate on Plato’s distinction between mimesis and diegesis, as two distinct modes of representation.

14  •  I ntroduction

or category but a scalar property that can be implemented to variable degrees. In a fuzzy-set conception of narrativity, the texts that Richardson describes as antimimetic occupy the margins. They are certainly fictional, and their transgressive nature with respect to prototypical narrativity indicates a literary ambition on the part of their author; but if the object of narratology is narrative texts, rather than artistic or fictional ones, narratology does not have to place the strangest creations of the postmodern imagination on the same footing as realistic novels, genre fiction, or fairy tales.8 Unnatural narratology must be commended for its exploration of experimental fiction (note that I write fiction, not narrative), but for an approach based on a binary opposition between natural and unnatural narrative, it has shown a disconcerting lack of interest in defining its polar opposite. The term natural narrative is often used by sociolinguists to describe conversational narrative of personal experience (as it is in Monika Fludernik’s seminal Towards a “Natural” Narratology [1996], which provided, through reaction, the inspiration for the movement), but it is obvious that literary narrative, being written rather than oral, predesigned rather than improvised, public rather than private, usually fictional rather than factual, will differ from conversational narrative through many features. While it may be true that faceto-face, oral storytelling is the ur-form of narrative, and that spreading news or gossip within the group may have been its original function, narrative has since diversified according to social function and to media, as human societies became more complex and developed various means of inscription and transmission for narrative data. All narratives are artificial, if by this term one understands constructed by the mind according to the demands of a certain social purpose rather than found in the world. Given the inability of unnatural narratology to define its opposite, it is no surprise that the movement has found unnatural features in virtually all of literature, especially in realistic narratives, the genre in opposition to which it originally defined itself: what 9 could be more unnatural than omniscient narration? The territory of unnatural narratology is continually expanding, and it will ultimately run natural narrative out of existence.10 8. Richardson (2021) reiterates his view that “antimimetic” texts are fully narrative in his response to a review by Gerald Prince of his book A Poetics of Plot for the Twenty-First Century: Theorizing Unruly Narratives. 9. Omniscient narration is, however, a systematization of a feature that occurs frequently in conversational narrative, and that is due to our propensity to use our “theory of mind” (Zunshine 2006): namely the representation of other people’s thoughts to explain their actions. 10. For a more detailed critique of unnatural narratology, see Ryan 2016, available on my website: http://marilaur.info/unnatural.pdf.

G rounding N arratology in the Concept of W orld  •  15

Of the four approaches featured in NT, cognitive narratology is the one I am most willing to be associated with because it is a very broad field with many ramifications, and one can pick and choose which ones to pursue. Yet I find Herman’s characterization of his approach in NT problematic. Herman does not label this approach cognitive narratology but rather the study of “the nexus of narrative and mind.” His reason for doing so (explained in NT 221) is that in philosophy of mind, cognitivism is associated with a conception of the mind as a disembodied computer that manipulates symbols. This conception is now widely rejected in favor of approaches such as enactivism that stress the inseparability of mind, body, and world. But insofar as narrative is the product of a mind addressing other minds, there is hardly any aspect of narrative that does not participate in this so-called nexus; we can therefore regard Herman’s formula as standing for the whole field of narratology. Then what is one doing when one does cognitive narratology? I have exposed in Ryan 201011 the dangers and limitations of some narrative applications of cognitive science: hard scientific studies based on brain scanning that cannot bridge the gap between neuronal activity and meaning formation, psychological experiments that test the reaction of subjects to simple texts that cannot represent the complexity of literary narratives, and overreliance on concepts imported from the more speculative disciplines of the cognitive sciences, such as social psychology or philosophy of mind, at the expense of trusting the ability of our own minds to figure out how we create, decode, and use stories.

Overview of the Book The arc of this book leads from the re-evaluation of some of the most foundational narratological concepts in the light of the conception of narratology that I have sketched above to the examination of narrative applications that highlight, expand, or even call into question the notion of storyworld. Chapter 1, “Truth: Discourse Types and Theories of Truth,” investigates various definitions of truth, both standard philosophical (correspondence, coherence, and consensus) and nontraditional (declarative), for the purpose of using this diversity of conceptions as a distinctive feature for a taxonomy of genres. The authoritative truth of myth, which tolerates no competition, and the truth-as-make-believe of fiction, which is protected from competing statements by the autonomy of the fictional world, are contrasted with the scientific conception of truth as correspondence, a conception which thrives 11. Available on my website: http://marilaur.info/cogninarr.pdf.

16  •  I ntroduction

in a competitive environment, since truth claims must pass the test of alternative representations. In chapter 2, “Fiction: The Possible Worlds Approach to Fiction and Its Rival Theories,” I compare and contrast five theories of fiction: (1) the naïve referential theory; (2) the invention-based, or rhetorical theory, proposed by Nielsen, Phelan, and Walsh; (3) Searle’s speech-act-based theory; (4) Walton’s make-believe theory; and (5) my own theory, based on the notion of possible worlds and inspired by David Lewis. These various theories are assessed in terms of the answers they give to a number of theoretical questions, with the ultimate goal of deciding whether the border between fact and fiction is gradual or binary. In the first part of chapter 3, “Narrator: Decomposing a Theoretical Primitive,” I compare standard conceptions of narrator as a mandatory component of language-based narratives with newly popular (or recently revived) theories regarding narrators as optional. In the second part, I argue that the notion of narrator is not the theoretical primitive for which it has been taken so far. “Narrating” a story is a complex activity which can, and should, be analyzed into distinct functions. Chapter 4, “Characters: Textual, Philosophical, and ‘World’ Approaches to Character Ontology,” compares three conceptions of fictional characters: 1. A “textualist” conception, represented by Roland Barthes, who regards characters as collections of semes, and insists on their subordination to the demands of plot. 2. A philosophical conception, which asks how statements referring to characters can receive a truth-value, and ascribes to characters the status of “abstract artifacts.” 3. A conception inspired by PW theory, which theorizes characters from the point of view of the storyworld. It will be argued that once one adopts an internal point of view, characters are imagined not as incomplete creatures made of language but as possible persons sharing the ontological completeness of the inhabitants of the real world. Have you ever encountered events in an otherwise reasonably well-plotted story that made you groan? It may be an overused device, a highly improbable coincidence that betrays the heavy hand of the author, a deus ex machina that resolves an otherwise hopeless entanglement, or a poorly motivated, outof-character behavior. In chapter 5, “Plot: Cheap Plot Tricks, Plot Holes, and Narrative Design,” I analyze a number of events that I regard as CPTs (cheap plot tricks) or as PHs (plot holes), focusing on their functionality for the story as a whole as well as on their prevalence in certain types of narrative and in certain historical periods. Plato conceived diegesis as narration, but he distinguished three ways to narrate: simple diegesis, where the poet speaks in his or her own name;

G rounding N arratology in the Concept of W orld  •  17

mimetic diegesis, where the poet imitates the speech of various characters; and mixed diegesis, which includes both modes. Classical narratology has largely ignored Plato’s distinction and excluded texts entirely based on the mimetic mode, but this exclusion has been challenged by the expansion of narratology to media such as theater, film, computer games, and comics, all of which combine the diegetic and the mimetic. In chapter 6, “Mimesis and Diegesis: Complementing Each Other,” I discuss the cognitive affordances and limitations of the two modes, and I explore how these limitations can be overcome by the use of mimetic elements in diegetic narration, and vice versa. Chapter 7, “Parallel Worlds: Physics, Narrative, and the Multiverse,” explores the idea that reality consists of a plurality of worlds. In physics, the existence of parallel worlds has been postulated on the cosmic level to describe what lies on the other side of black holes, and on the level of subatomic particles to avoid the paradoxes of quantum mechanics. In fiction, the idea of a plurality of words is represented by standard themes such as trans­ world exploration, alternate history, and time travel, as well as by narratives explicitly inspired by the many-worlds interpretation of theoretical physics. In this chapter I explore several of these narratives, focusing on the notion of counterpart relations between individuals who inhabit different worlds, as well as on the plot possibilities created by these relations. Chapter 8, “Impossible Worlds: Dealing with Logical Contradiction,” approaches the problem of aesthetic illusion through the examination of a category of texts that inhibits this experience: texts that create logically impossible worlds. The notion of impossibility is contrasted with the notion of the unnatural, which describes worlds that differ from ours but remain accessible to the imagination. Four types of impossibility are described: ontological impossibility (i.e., metalepsis and co-presence in the same world of characters originating in different texts), impossible space, impossible time, and impossible texts. These texts require the ability to shift back and forth between a narrativist and a textualist perspective, so as to appreciate the text both as a deliberately failed world representation and as a virtuoso verbal performance that pushes back the limits of the textually (im)possible. When the big leap forward of computer technology took place for the general public in the 1980s and ’90s, virtual reality technology (VR) was touted as the next big thing that digital media would bring into our lives. But VR did not live up to its expectations, and after the year 2000, it faded from the radar of popular interest. After a period of hibernation, VR has recently regained attention as a form of entertainment, and “immersive experiences” are now proliferating. Currently available VR narratives are distinguished from other digital narratives through three-dimensional images, interactive panoramic

18  •  I ntroduction

representations, and the ability to manipulate our experience of our own body. In chapter 9, “Virtual Worlds: Narrativity and VR Technology,” I discuss several projects that use some of these resources. Basing my judgment on this limited corpus, I assess the potential of VR narratives with respect to four kinds of immersion: ludic, spatial, temporal, and emotional. The technological innovations of the past thirty years have facilitated the development of cult narratives that expand in many different media and create communitie of dedicated fans who discuss these narratives online or expand their storyworlds through their own contributions. The discourse of the entertainment industry wants us to believe that, thanks to technology, storytelling will never be the same. In chapter 10, “Transmedia Worlds: Industry Buzzword or New Narrative Experience?” I examine this claim critically by distinguishing, on one hand, bottom-up manifestations of transmediality that exploit the popularity of an already established monomedial narrative, growing in a random, uncontrolled manner through a snowball effect, and, on the other hand, the very rare case of top-down projects conceived from the beginning as developing a storyworld across many media. Then I examine the discourse of the industry through the rhetoric of how-to books about constructing transmedia worlds, and I conclude by sketching what narratology can do about transmedia storytelling beyond acknowledging its existence.

CHAPTER 1

Truth Discourse Types and Theories of Truth

In the late 1990s, when the original version of this chapter was written, the dominant academic philosophies of postmodernism and deconstruction were busy denouncing the notion of truth by regarding it as an expression of political, social, and cultural assumptions, as a “narrative” created by “power.” Belief in the possibility of a truthful, objective representation of the world was denounced as “naïve realism” and attributed to a lack of intellectual sophistication. Now that we have supposedly entered a post-truth era, the tide has turned and the two sides have exchanged positions. The people who believe in a “deep state” or in conspiracy theories such as QAnon, and who reject science by denying climate change or resisting vaccination, are accused of living in an alternate reality. It is now the intellectuals who pose as defenders of truth and insist on fact-checking the claims of politicians. While the deconstruction of truth was a left-wing phenomenon, post-truth is a right-wing one, and it certainly lacks the sophistication with which philosophers such as Derrida or Foucault analyzed the notion of truth. It would therefore be unfair to regard the post-truth phenomenon as the triumph of deconstruction; if anything, it results from its perversion. But as Lee McIntyre writes, “Postmodernists have contributed to [the post-truth phenomenon] by retreating within the subtlety of their ideas, then [are] shocked when they are used for purposes outside of what they would approve” (2018, 127).

19

20  •  C hapter 1

If there is one event that embodies the spirit of the postmodernist assault on truth, this event is the Sokal hoax. In 1996 NYU physicist Alan Sokal sent an article titled “Transgressive Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity” to the journal Social Text, a champion of postmodern theory. Laced with absurdities, such as claiming that pi is a variable, the article argued that recent developments in science and mathematics, such as Einstein’s theory of relativity, Heisenberg’s principle of uncertainty, Gödel’s proof of the incompleteness of axiomatic systems, chaos theory, and the mathematical investigation of nonlinear phenomena have (scientifically?) demonstrated that physical “reality,” no less than social “reality,” is at bottom a social and linguistic construct; that scientific “knowledge,” far from being objective, reflects and encodes the dominant ideologies and power relations of the culture that produced it; that the truth-claims of science are inherently theoryladen and self-referential. (Sokal [1996] 2000b, 11–12)

In an article later published in Lingua Franca ([1996] 2000a), Sokal revealed that the essay was a fabrication meant to challenge the rigor of the postmodernist critique of science. He believed that he could get a text with the most nonsensical scientific claims published in Social Text if it conformed to the position of the journal. During a public discussion with the editors of Social Text, Sokal emphasized the incompatibility between the belief of scientists that “Native Americans came to North America across the Bering Strait more than ten thousand years ago and the position of Indians who believed that their ancestors rose, fully formed, from a subterranean spirit world” (reported in Gladwell 1996, 27). Both theories, he claimed, cannot be true at the same time. The choice of example was unfortunate because it exposed Sokal to accusations of ethnocentrism, which his opponents were only too happy to take up. Since myth is taken here to represent Native American culture and science Western civilization, choosing one version over the other implies the rejection of an entire system of cultural beliefs. Though they represent diametrically opposite positions, Sokal and his opponents had at least one point in common: both subjected myth and science to a unique concept of truth and a unique mode of evaluation. Sokal reads myth as if it were faulty science, while the postmodernists read all texts as world-constructing fictions. Sokal wants the text to reflect “the order of things,” a criterion that leads him to proclaim the superiority of science over myth on the basis of its reliance on experimental verification, while the postmodernists hold that the truth-value of the text is relative to a language, or

T ruth  •  21

cultural tradition, or conceptual scheme, and that all conceptual schemes are equally valid and equally relative. This position supports the attractive ideas of plural perspectives and of a nonhierarchical system of discourse genres, but the price to pay for epistemological tolerance is the trivialization of the notions of truth and validity, since under this model every conceptual scheme generates its own truth and falsity becomes problematic. Both sides fail to take into consideration the alternative that I propose to investigate in this chapter: that there are multiple theories of truth, and that we may apply different truth criteria to different types of texts.1 Some people may indeed accept Darwin’s theory of evolution while regarding the Bible as divine revelation, without being bothered by the contradiction. What I propose to do in this chapter, then, is to use different conceptions of truth and validity as a basis for distinguishing types of discourse. Philosophy has proposed a variety of answers to the question of the nature of truth: correspondence theories and their close relatives, formal semantic theories (epitomized by Tarski’s formula “‘snow is white’ is T iff snow is white”); theories regarding truth as a matter of coherence with other truth-bearers; and consensus theories making truth dependent on what passes as true. The spirit of the doctrine of truth as correspondence is captured by formulas such as “A sentence is true if it designates an existing state of affairs” (Alfred Tarski, The Semantic Conception of Truth)2 or “To say of what is that it is not, or of what is not that it is, is false, while to say of what is that it is, or of what is not that it is not, is true” (Aristotle, Metaphysics).3 In a correspondence theory, truth is a property attached to certain representations independently of opinion, or even of the actual occurrence or formation of these representations in somebody’s mind. The theory supports the claim that there are unknown truths, because there are true propositions waiting to be asserted, and true representations waiting to be discovered. The coherence theory is more equivocal on this point, but although it makes truth dependent on the rules of logic (arguably universal) and man-made conventions (highly genre- and culture 1. Does this position amount to relativism? I don’t think so: each theory relies on definable criteria of truth. Relativism would be a generalization of truth as consensus to all areas of knowledge, a rejection of the possibility of rigorous truth criteria, or a refusal to distinguish beliefs from truth. 2. Quoted by Kirkham 1992, 70. The formula appears several times in Tarski’s writings, for instance in “Truth and Proof,” p. 63. 3. Quoted by Kirkham 1992, 170. Kirkham himself offers the following formal definition of the correspondence theory: (t) {t is true iff (∃x)[(tRx) and (x obtains)]}

where t is a truth-bearer (i.e., an expression or belief that may or may not be true), x a state of affairs, and R a relation of representation or standing for.

22  •  C hapter 1

dependent), it still allows the possibility of truth being discovered rather than made: once the rules are established, it is the rules, and not our opinion, that determine the truth of statements. This possibility disappears in consensus, or pragmatist theories. If, as Richard Rorty claims, “there is nothing to be said about truth save that each of us will commend as true these beliefs that he or she will finds good to believe” (1989, 3), judgments of truth are declarative acts, and truths are made by these judgments. Sokal’s confidence in the epistemological superiority of science over myth expresses the traditional allegiance of the field he represents to the correspondence theory. The belief in the possibility of satisfying this theory—the most demanding of the three sketched above—is the apple of discord that separates him and the so-called Enlightenment tradition from postmodern doctrine. For the correspondence theory to yield judgments of truth and falsity, it must be supported by an appropriate metaphysical stance: reality must not only exist independently of the human mind but also be readable by the mind. In general, postmodernists do not deny the existence of an autonomous physical world; the skepticism that leads postmodernism to reject the correspondence theory of truth is much more concerned with the knowability of the real than with its objective existence. This skepticism is supported by the following arguments.

Linguistic Relativism One of the firmest beliefs of postmodern theory is that Saussurian linguistics constitutes the ultimate authority on the nature of language. Equally unchallenged among literary critics of this persuasion is the view that the insistence by Saussure on the arbitrary nature of language has dealt a fatal blow to the idea that language can speak objective truth about the world. Saussure, of course, was not the first to observe that the relation between meaning and sound is based on convention rather than naturally dictated. But his formulation of the principle of arbitrariness is more radical than the view of his predecessors because it affects not only the relation between signifier and signified but also the delimitations created by language in the very substance of sound and meaning. For Saussure, signs acquire their meaning not from vertical relations to objects in the worlds but from horizontal relations with other signs: “In language there are only differences without positive terms” (1966, 120; italics original). Language is not a repertory of names for inherently distinct objects but a self-enclosed, self-regulating system whose terms receive their value from their relation to all the other terms of the system.

T ruth  •  23

Saussure’s famous example of the role of the system in creating values compares French mouton with English sheep. The two terms have different values because in English sheep contrasts with mutton, while in French the term covers both animal and meat. The conceptual system of the French thus involves one entity where the English see two, but the only thing we can say about these entities, if we remain true to the differential principle, is that their value occupies the territory left unclaimed by their neighbors. Now if every language cuts up semantic substance in its own way, rather than reflecting a naturally given order of things, how can verbal representations entertain a relation of resemblance to the world, how can they, in other terms, correspond to reality? For the proponents of deconstruction, as Robert Scholes points out, the answer leaves little doubt: if language is arbitrary, the correspondence theory is wrong: Many believing deconstructivists .  .  . accept as an article of faith that it is impossible to utter the truth because there is an unbridgeable gap between human language and the world. According to this view, everything we speak or write is fictional because there is no such thing as the literal meaning of words . . . Because perception is always coded by language, and because language is never “literal” but always figurative, perception is never accurate, is always distorted. (1993, 179–81)

This argument rests on the assumption that we can only make true statements about the world if the structure of language, prior to its use, truly reflects the structure of the world. But according to logicians, the predicate true applies to individual utterances and expressions, not to language as a code. Sentences make truth claims, but individual words do not. Confusing language and language use, langue and parole, this line of reasoning tells us that in a rather circular way, language in itself must be a true statement about reality if it is to provide the tools for making true statements about reality. Only if language were naturally given, spoken by the world about the world, would it be possible for its users to speak truly about the world. But as J.  L. Austin observes, “there is no need whatsoever for the words used in making a true statement to ‘mirror’ in any way, however indirect, any feature of the situation or event” (1970, 125). Given the proper semantic conventions, a oneword expression could represent the fact that snow is white just as well as a three-word sentence. Correspondence is not mimesis but simply correlation between the meaning conventionally encoded in an expression and a state of affairs obtaining in the world. It is obvious that in another language where snow means “coal” the sentence “snow is white” would be false—but even in

24  •  C hapter 1

this language, the truth-value of that language’s way of saying “snow is white” would arguably be determined by comparison with an external state of affairs. Another questionable assumption of the Saussure-inspired critique of truth as correspondence lies in the claim that when using language in a concrete situation, we deal with the full value of words. But it is hard to see how something meaningful (let alone true!) could be said in a sentence predicating a property such as the full Saussurian value of mortal (both “fatal” and “subject to death”) of any entity. Socrates is mortal will activate one meaning, and this wound is mortal the other. There is no compelling evidence that in areas other than poetry, where polysemy can be deliberately cultivated, purely differentially defined values have psychological reality and communicative value: a child or foreign speaker who learns only one of the meanings of mortal is perfectly able to put this meaning to use. If our experience of the world enables us to take apart what language bundles together, the somewhat arbitrary semantic bundles covered by signifiers must not determine as rigidly as postmodernists believe the categories of thought and the structure of perception.4

Postmodernist Perspectives on Truth There is a tendency among traditional philosophers and postmodernists alike to assume that a valid truth theory must apply equally successfully to all assertive utterances and to all types of semantic contents, and to reject for that reason any relative or limited concept of truth. In practice, however, philosophers tend to concentrate on one type of statement and declare it exemplary: for the proponents of correspondence theories, for instance, the paradigmatic statements typically deal with material objects and sensory properties. Understandably, a theory of truth tailor-made for “snow is white” will be vastly different from a model meant to assess the truth of a biography, a work of history, or a psychoanalytical analysis. The postmodernist alternatives to the notion of truth as correspondence are no less focused on certain types of statements than the theory they want to dispel. For Barbara Herrnstein Smith, the paradigmatic case is value judgments: “Madame Bovary is a great work,” or “Marilyn Monroe was beautiful.” It is obvious that the state of the world cannot decide the accuracy of such statements. Even though they take the form of assertions, value judgments 4. Among linguists, belief in the Sapir-Whorf hypothesis (according to which thought is crucially determined by linguistic categories) has strongly waned since Chomsky’s pronouncements on the universal and wired-in nature of linguistic competence.

T ruth  •  25

must be assessed in terms of felicity of utterance or in terms of the speaker’s commitment: the speakers of “Marilyn Monroe was beautiful” make the statement felicitous if they sincerely believe this to be the case. Assuming that the concept of truth must be a matter of correspondence if it matters at all, Smith argues that because value judgments do not fulfill the correspondence requirements, the whole idea of truth-value must be rejected for any kind of statement: The dominant tradition in post-Kantian aesthetic axiology has characteristically offered to demonstrate that such judgments do have truth-value . . . We are, however, approaching the issue from a different—in fact, reverse— direction, the procedure and objective here being not to demonstrate that value judgments have as much claim to truth-value as factual or descriptive statements but, rather, to suggest that, just as value judgment do not have but also do not need truth-value in the traditional sense, neither, it seems, do any of those other forms of discourse. (1988, 99–100)

Richard Rorty’s position is less radical than Smith’s: his ambition is to do away not with the notion of truth altogether but only with the correspondence theory. While he recognizes that with certain types of statements “the world decides the competition between alternatives” (example: “red wins” versus “black wins” [1989, 6]), he points out the insufficiency of this criterion for statements about fictional entities, such as “Sherlock Holmes lived on Baker Street.” (He means the sentence as a comment of the reader, not as an authorial or narratorial speech act.) In this case there is nothing “out there” in the world that makes the statement true or false, yet all the people familiar with Conan Doyle’s stories agree on the truth of this statement. Since for many philosophers reference and correspondence presuppose the existence of material entities, Rorty claims that we have two alternatives to deal with Sherlock Holmes statements: “a ‘pure’ language-game approach which dispenses with [reference and correspondence] altogether, or a rigidly physicalist approach which interprets them in terms of physical causality” (1982, 127). While the physicalist approach (the position toward which Bertrand Russell was leaning) denies truth-value to all statements that do not refer to material objects, the language-­game approach “separates semantics from epistemology so drastically that semantics will have no interesting distinctions to make between truth about fact and about fiction” (1982, 127). The kind of language game proposed by Rorty is a form of consensus theory according to which truth is a matter of “warranted assertibility.” All it takes to make a statement true is people finding motivations to believe it, and all there is to say about truth is that it

26  •  C hapter 1

is “made” and not “found”: Rorty sees no need to probe further into the various reasons that lead members of a culture to accept “Sherlock Holmes lived on Baker Street” and “The earth revolves around the sun.” That the theory is unable to differentiate these two cases is no loss for Rorty, because he regards “the need to distinguish sharply between science and poetry” as an exclusively Western preoccupation (1982, 132), and his avowed goal is to liberate philosophy from the “scientism” of Western tradition: Getting rid of the correspondence theory of truth is not a discovery about the nature of a preexistent entity called “philosophy” or “truth.” It is changing the way we talk, and thereby changing what we want to do and what we think we are. (1989, 20)

By proposing fictional sentences as the basis for a theory of truth, Rorty hopes to put philosophy in a new key, one in which language will be acknowledged as a creative force, poets will be honored as “the vanguard of the species,” and literary criticism will become the “presiding intellectual discipline.” (Here I quote a paraphrase of Rorty proposed by Peter Lamarque and Stein Haugom Olsen [1994, 194]).

Diversifying the Concepts of Truth The postmodern critique may not have dealt a fatal blow to the concept of truth, but by showing the limitations of the theory of correspondence it invites us to reconsider the wisdom of constructing a general theory of truth on a single type of statement. Speech act theory has shown that different types of sentences require different appropriateness conditions; truth is only relevant for 5 assertions and, to some extent, for declaratives (those utterances that, according to J. L. Austin, are made true by virtue of the authority of the speaker, such as “Out!” being yelled by a baseball umpire). It doesn’t take training in formal semantics to realize that different types of assertions are true under different conditions. While “snow is white” or “Napoleon died in exile on Saint Helena” are made true by the way the world is, an analytical statement such as “bachelors are unmarried” is true independently of any external state of affairs, “I love you” depends on the speaker’s mental state and transparency to herself, “if a batter bunts foul with two strikes against him, he is out” presupposes an 5. What I call here declaratives is a category also known in speech act theory as performatives. I do not follow the usage of those linguists and philosophers who use declaration as a synonym for assertion.

T ruth  •  27

original pact, or consensus among players, and given the premises “all men are mortal” and “Socrates is a man,” we can conclude that Socrates is mortal without knowing anything about the world: truth in this case is a matter of coherence (given the rules of logic) with the two premises. If correspondence, coherence, consensus, analytical, and declarative (performative) conceptions of truth are all needed on the sentential level, it will not take an unnecessary cluttering of the conceptual apparatus of semantics to apply them to the global textual level. Shown below is an attempt to ascribe truth, or acceptability conditions, to various types of discourse. The list includes both standard technical concepts of truth and concepts which philosophers would not regard as such because their vehicles are protected from falsity (the declarative types). It would exceed the scope of this chapter to propose a detailed analysis of each of these categories. In the discussion to follow I propose to concentrate on the categories relevant to science, myth, and fiction, the genres targeted by the Sokal controversy. Correspondence: Natural sciences, social sciences, history, traditional journalism Coherence: Mathematics and logic Analytical: Dictionaries Speaker’s belief (intuitive): Mystical writings Declarative by way of consensus: Laws, game rules Declarative by way of invitation to make-believe: Fiction Declarative by way of revelation: Myth Hybrid, instable, or undecidable: Philosophy, literary criticism, “theory” Figural: A type of truth achieved in addition to one of the above, arguably by literature and for some communities by myth

In contrast to the postmodernist attempts to reduce all texts to fiction, this model does not rely on metaphysical assumptions concerning the relations between language, the world, and the mind. In classifying certain modes of discourse under the rubric correspondence, for instance, it does not claim that texts can accurately reflect reality but only that correspondence forms the publicly acknowledged goal and criterion of validity of the genre in question. Nor will it matter whether politics, such as the pursuit of personal interests, sullies this ideal or not. Whereas postmodernism takes a “we know better” attitude toward certain genres, regarding their practitioners as fundamentally deluded and in urgent need of education, I consider the culturally approved patterns of behavior associated with different types of text as constitutive of these cat-

28  •  C hapter 1

egories, and as an integral part of our generic competence. My approach can be described as a pragmatism of the intentional level: it is concerned neither with what philosophers think we can do with language nor with the complex, partly subconscious motivations that determine what we actually do, but with the goals we admit to ourselves.

Correspondence Genres In the correspondence group, texts are proposed as representations of a physical reality external to both language and the mind. According to at least one version of possible worlds theory, the only world that fulfills these requirements is the actual world.6 All correspondence texts consequently share the same reference world. Since they project versions of a common reality, their representations are fundamentally commensurable, and they entertain a competitive relation with other texts of the same category: the reader decides to accept or reject them on the basis of other sources of knowledge concerning the same world. The polemical, tentative character of their claims is most strikingly exemplified by the case of science. The postmodernist critique of correspondence aims at dispelling the aura of epistemological supremacy that contemporary society tends to project on science, but the practice of the scientific community tells an entirely different story: far from speaking from authority, scientific discourse is subjected to a rigorous process of challenge, testing, and discussion, and its claims are never safe from future discoveries, theoretical revolutions, and alternative proposals. Like all correspondence genres, science operates under the principle that its representations must be updated and its models sometimes discarded if they are unable to account for new evidence. If a complete human skeleton were found fossilized in a preJurassic layer of the Grand Canyon, for instance, evolution would have to be thoroughly rewritten, and the theory that would replace it would come much closer to the Native American myth discussed by Sokal. Myth, by contrast, would never be open to alternative explanations. To say that a text as a whole must be evaluated in terms of correspondence with reality does not mean that every one of its component statements falls under this mode of evaluation. Discourse types such as science and history consist not only of descriptions that make direct claims about the world (the data) but also of statements that interpret or generalize these claims. The for

6. The alternative version is David Lewis’s modal realism, discussed in the introduction.

T ruth  •  29

mer are evaluated in terms of their accuracy with respect to the facts, the latter on the basis of their coherence with respect to the data. Moreover, science relies heavily on mathematics, a discipline grounded in the coherence mode. The importance of the coherence factor is a function of the subject matter and level of abstraction: the deeper the text probes into the mental, the general, and the conceptual, the more prominent will be the role of statements derived through inferences, which means by principles of coherence.

Fictional Truth I will be brief in this section since the notion of fiction is discussed in great detail in chapter 2. Here I adopt a world-based, make-believe approach inspired by the work of Lewis (1978) and Walton (1990) which analyzes fiction as a layered act of communication. On one level, a real-world author addresses a text to an audience located in the same world; the text is taken by the reader to be the representation of an imaginary world created through the very process of representing it. On another level, a narrator located in what is from the viewpoint of author and reader an alternative possible world presents a report that passes as a representation of an autonomous reality.7 Insofar as they claim (or mimic) reference to this reality, the narrator’s declarations are potentially true or false, but since the text is the only source of information about the fictional world, sorting out the true from the false is impossible. If readers are to construct a mental image of the fictional world, they must therefore take the narrator’s declarations at face value. (An exception must be made in the case of unreliable narration, but even unreliable narrators must be believed most of the time for the inconsistencies in their reports to be detected.) The author instructs the reader to regard certain statements as true in the fictional world, and these “fictional truths,” to use Walton’s term, cannot be contested without violating the conventions of the genre. Insofar as the fictional text constitutes the only mode of access to its reference world, fictional discourse is removed from the law of competition to which factual texts are subjected. Once a fictional truth is established, no other text can challenge it. A new version of an old plot, a sequel, or a retelling does not invalidate the original version of the storyworld but either expands it, or creates a closely related storyworld. 7. As we will see in chapter 3, narrator-optional theories of fiction deny the existence of this level in third-person, heterodiegetic narration, but would accept it for first-person homodiegetic.

30  •  C hapter 1

Literary Truth This analysis of fiction does not attempt to capture the aesthetic or literary dimension of the text. If we take literature in a canonical sense, meaning “the most honored pleasure-texts of a culture,” the category only overlaps with fiction: there are nonfictional literary texts (well-written autobiographies), just as there are nonliterary fictional texts (e.g., genre fiction). What I call here fictional truths are not powerful ideas, inspiring revelations, bold insights, or nutritious food for thought, but simply the “facts” that must be accepted to construct a personal representation of a fictional world. “Charles Bovary is a country doctor” is a fictional truth because it has axiomatic value for the world of Madame Bovary, but I would hesitate to label the first sentence of Anna Karenina, “Happy families are all alike; every unhappy family is unhappy in its own way,” a fictional truth, because the reader may reject this statement and still manage to be caught up in the story. (Opinion statements are not authoritative.) As this example suggests, the vast majority of fictional truths are not deep, meaningful messages that can be exported to the real world, and the meaningful messages conveyed by fiction, if any, are not necessarily fictional truths, because they do not have to be explicitly stated. Since the notion of fictional truth has nothing to do with significance and aesthetic value, many critics feel a need to complement it with a notion of literary truth. While all fictions convey fictional truth, only some of them, arguably the best, reach the level of literary truth.8 Definitions of literary truth typically invoke the general, the abstract, the spiritual, and the existential. Suitable subject matters are Love, Death, the Meaning of Life, the Struggle of Good against Evil in the Human Soul, Moral Responsibility, What It Means to Be Human, and—especially prominent in postmodernism—the Nature of Language, Truth, Knowledge, and Art. In many respects, the two types of truth are polar opposites: fictional truth concerns the storyworld, but literary truth is transportable to the real world; fictional truth is about the actions and properties of characters, but literary truth is a knowledge that (hopefully) enriches readers and guides them in life; fictional truth is literal, but literary truth is figural (as Nelson Goodman writes: “Don Quixote, taken literally, applies to no one, but taken figuratively, applies to many of us” [1978, 104]); fictional truth is a relatively straightforward technical concept, but literary truth is so elusive an idea that it can break the language of those who try to capture its 8. A stand against the theoretical importance and contribution to aesthetic value of literary truth is taken by Lamarque and Olsen (1994). For a skeptical view of the ability of literary fiction to convey truths for the real world, see Currie (2020). I personally wish to leave open the question of whether literary truth is a necessary component of aesthetic value.

T ruth  •  31

essence. To wit the rather clumsy sentence by Colin Falck: “But what makes a story—or any literature which is not in the form of a story—significant is that it gives us insights into, or that it reveals or discloses, something of importance about what human life, or therefore reality as it is humanly experienced, is really like” (1994, 108).

Mythical Truth Nineteenth-century positivism called myth a superstition. Fiction never suffered the insult. In the language game of fiction, the false is regarded as true, but the player knows that it is only a game. In myth, which is not a game, the false is seriously regarded as true, and the believers are deluded. This at least was the position of a culture that had banished myth from its textual system and viewed it with condescendence as the manifestation of a primitive stage in the evolution of thought. But if we approach myth from the point of view of those who actively engage in this mode of discourse, it not only speaks the truth, it does so necessarily. Myth is the voice of Truth itself, the foundation of a culture, a definitive representation that refuses to acknowledge the existence of any other version. Its truth is so secure that the narrators of myth never need to justify the source of their information (Booth 1996, 244), and its authority so complete that “belief in the myths of a community is compulsory” (Pavel 1986a, 61). Since the truth of myth is guaranteed by its origin—ancestors, tradition, or the gods—myth shares the declarative status and world-creating power of fictional discourse. But the analogy stops at the world of reference: while fiction establishes its truth for a world that is ontologically not ours (we must transport ourselves in imagination to regard it as actual), myth concerns the most real of all worlds, the very center of the believer’s ontological system. Mircea Eliade (1975) observed that the heroes of myth live in illo tempore, in a sacred time and place rigidly delimited from the realm of the profane, but this separate domain is an integral part of the real rather than one possible world among many others. From this derives what is perhaps the main difference between myth and fiction: there is only one corpus of myth per culture because there is only one actual world, but there are many fictions, living in peace with each other, because each one creates a different possible world. As a representation of the real, then, myth targets the same reference world as scientific discourse. Insofar as they attempt to provide explanations for natural phenomena, the two modes of knowledge also overlap in their subject matter: science tells us what optical phenomena cause rainbows, myth

32  •  C hapter 1

tells us how rainbows were created as a sign of an alliance between man and God; science explains menstrual periods in terms of hormonal cycles, myth may do so in terms of a mystical bond between women and a moon goddess. The solutions, of course, will be as different as experimentation and authority, the respective justification for the truth claims. Whereas science tries to detach the object of study from the observer in order to reach objectivity (whether or not this is possible), myth integrates the subject into a cosmic order. Its truths, therefore, have the human significance that critics claim for literary truths. While the subject matter of science is the physical world, myth is more properly concerned with the interaction of the physical and the spiritual (Falck 1994, 117). As the raging debate between creationists and partisans of evolutionary theory in contemporary American society reminds us, the common reference world and the overlapping concerns of myth and science place the two modes of discourse in an inherently competitive relation: if one is true, the other cannot be. Having to contend with other representations is a normal situation for science, but in the case of myth, it is a sign of degeneration. The three types of discourse I have discussed above adopt strikingly different stances toward competition: science thrives under it, fiction protects itself from it by isolating itself in its own private world, myth cannot accept it, because it is threatened in its very essence by any challenge to its authority. The rivalry of science and myth that divides modern cultures internally and pits them against each other was never supposed to happen: the narratives that are still active as myths in contemporary society are mostly the legacy of an age when science was still in its prehistory. If the conflict between myth and science is to be resolved, the mode of belief in myth must be altered. There are four ways in which mythical claims can be taken: (1) as authoritative, inspired, community-defining, literal truths (the fundamentalist mode); (2) as authoritative, inspired, communitydefining but potentially figural truths (the mainstream religion mode); (3) as literary truths, offering one source of inspiration among many others; (4) as fictional truths, forming entertaining narratives about imaginary worlds. Of these four modes of reading, the first two retain the conception of truth that I propose here as constitutive of myth, while the last two effectively kill myth but keep its memory alive by recycling its narrative material into another genre (as discussed in Pavel 1986a). The subgroups of contemporary Western cultures that still have myth in their active system of genres must decide whether to accept the fundamentalist view and reject science altogether; start an “alternative science” to support their literal belief in myth; accept both myth and science but establish a hierarchy when the two come into conflict; or adapt myth to science by settling for the figural interpretation.

T ruth  •  33

Conclusion Neither entirely positivist nor relativist, the present attempt to pair discourse types with truth conditions advocates a pluralism that legitimizes positivism within a certain area but relativizes its pronouncements by acknowledging the validity of other concepts of truth within other domains. Diversifying truth and recognizing the existence of various language games does not entail an adherence to the view—pace Rorty—that all truth is language-made: among the games I describe, there are at least a few in which truth is not conceived as the product of conventions, agreement, or authority. One wonders what would happen if the postmodern critique of truth as correspondence succeeded in bringing the scientific community down to its knees, and made it abjure its faith in the possibility of reaching some degree of objective knowledge about the world. Would it help to pave the way toward a better world and a better science, or would it justify such behaviors as the worship of authorities, the cynical pursuit of personal interests, and a free play with data? A critique of science can only be successful if it promotes self-criticism in the scientific community, but the need for criticism would be thwarted without the competitive environment created by the correspondence model. Rather than telling the scientific community “your goal is illusory, give it up,” a constructive critique should explore to what extent science is faithful to its goal, and what obstacles stand in the way. As long as theory denies any validity to the concept of truth as correspondence, it will remain without impact on the practice of science, and it will be complicit in the rise of the post-truth era.

CHAPTER 2

Fiction The Possible Worlds Approach to Fiction and Its Rival Theories

The question of fictionality has long been neglected by narratology. The discipline developed as the study of literary narrative fiction, and the question of fictionality was either taken for granted or considered irrelevant to the project of exploring the many forms of narrative discourse on the grounds that since fictional narration presents a broader repertory of narrative strategies than the factual kind, any device that appears in nonfiction can arguably be imitated in fiction, while the reverse does not hold true. It was left to philosophers working in the Anglo-American analytic tradition, such as Searle (1975), Lewis (1978), and Walton (1990), to raise the question of what authors are doing when they are writing novels rather than claiming truth. Literary critics inspired by possible worlds theory took notice of this work and developed interest in the issue of fictionality (Eco 1979, Pavel 1986a, Doležel 1998, and myself [1991]). In the 1990s Genette joined the trend with his book Fiction and Diction (1991), while a philosophical tradition independent of possible worlds theory continued to develop with the work of Lamarque and Olsen (1994), Gregory Currie (1990, 2010, 2020), and Amie Thomasson (1999), to name only a few. In 2015 a theory of fictionality which the authors call rhetorical was introduced with great fanfare by narratologists Henrik Skov Nielsen, James Phelan, and Richard Walsh, and the notion of fictionality has played a prominent role in narratology ever since. In this chapter, I propose to compare and contrast five theories of fiction: first the so-called naïve theory, which 34

F iction  •  35

corresponds to the informal, or everyday use of the word fiction; then the pioneering theories of John Searle (1975), Kendall Walton (1990, but based on earlier articles), and David Lewis (1978), on which my own theory is based; and finally the rhetorical theory of Nielsen, Phelan, and Walsh. 1 These various theories will be assessed in terms of the set of phenomena they accept as fiction, and in terms of the answer they provide to a number of key questions, which I will define later.

The Five Theories In the naïve theory, fictionality means falsity. True statements refer to something in the world, while fictional statements do not refer to anything. It is in this sense that Donald Trump labeled Fear, a book by Bob Woodward that depicts the dysfunctionality of the Trump White House, “a work of fiction,” or that Frank Bruni wrote in the New York Times in 2019, “All presidents want to rack up triumphs that make them look and feel large. But none in my lifetime has spun so many fictions [as Trump] in the service of that.” The rhetorical approach2 shares with the naïve conception the association of fiction with invention, as opposed to truth, but it avoids the conflation of lies, errors, and deceit with fiction by insisting that fiction relies on an overt display of invention. The sender of fiction wants the audience to recognize the fictionality of the act of communication. The purpose of the proposal is to free the concept of fictionality from what the authors call generic fiction—literary forms such as novels and short stories, as well as fiction films—and to outline instead a theory that encompasses a wide variety of utterances: not only literary narratives but also “hypotheticals, counterfactuals, speculations, and other deviations from the actual” (Nielsen et al. 2015a, 64). According to the authors, the faculty of invention manifests itself in many genres of discourse: in informal conversation, through jokes and “kidding around,” in political speeches, through projections; in sermons, through parables; and in philosophy, through thought experiments. Fictionality may therefore appear locally in a globally nonfictional discourse, just as noninvented facts may appear locally in a global fiction, such as a novel. As an example of nonliterary fictionality, 1. My choice of theories is by necessity limited: it would take a whole book to broaden the discussion to a representative sample of thinking about fiction. Notably absent from consideration is recent work in France (Schaeffer 1999, Lavocat 2016) and Germany (Zipfel 2001, the articles in Klauk and Köppe 2014a, Rajewsky 2020). 2. I base my presentation of the rhetorical approach on the original articles (2015a and 2015b). The authors may have diverged from it and from each other in their later publications.

36  •  C hapter 2

the authors mention Barack Obama’s claim that Mitt Romney, his opponent in the 2012 presidential election, suffers from “Romneysia,” a condition obviously invented by Obama. Another, more controversial, example is Martin Luther King’s “I have a dream” speech, which does not describe an actual state of affairs but rather expresses King’s hopes for a future of equality between the races. To regard this speech as fictional, one must assume that since the future does not contain verifiable facts (only the past does), any discourse referring to the future constitutes an invention. An important point of the rhetorical theory is that fictional invention is never an end in itself but is a means to an end, a way to say something about reality. The authors view fictionality as a rhetorical device that people “use” in order to make statements about the real world: “Fictive discourse is not ultimately a means of constructing scenarios that are cut off from the actual world, but rather a means of negotiating an engagement with that world” (Nielsen et al. 2015a, 63). For instance, through the invention of Romneysia, a play on amnesia, Obama wanted to remind his audience that when Mitt Romney was governor of Massachusetts, he installed a universal insurance plan, which he later rejected when he ran for US president. Or, to take a literary example from the article, the young adult novel The Hunger Games, by Suzanne Collins, provides its readers a model they can emulate by showing how the heroine refuses either/or situations where both options are ethically unacceptable and finds instead a creative way to avoid the choices imposed on her. Moving back in time, let’s turn to John Searle’s conception of fiction as pretended speech acts (1975), which is the first attempt by a philosopher to define fiction. Searle begins his argument by comparing the beginning of a news report from the New York Times and the beginning of The Red and the Green, a novel by Iris Murdoch. He observes that both texts rely on the speech act of assertion, and that both follow the same lexical and grammatical conventions: consequently, there is no special language of fiction.3 Searle also rejects the idea that there is a special speech act for fiction, such as “writing a novel,” because novels are made of the same kind of speech acts (mostly assertions) as nonfictional narratives, such as history and news. What is it, then, that makes the novelist’s utterance act different from the journalist’s? In Searle’s speech act theory, every kind of speech act is regulated by specific semantic and pragmatic rules. The most important of the four rules of asser 3. Literary critics such as Dorrit Cohn (1999) have argued for “signposts of fictionality,” i.e., devices that mark a text as fiction, but when these signposts are stylistic devices (such as internal focalization) they are not constitutive of fiction, since a text can be a fiction without presenting them. On the other hand, when they are pragmatic (such as: in fiction, the narrator is not the author), they may be constitutive, but they are not really signposts, since they do not correspond to visible textual properties. The assumption that the narrator is not the author follows from the identification of the text as fiction, rather than leading to this identification.

F iction  •  37

tion is the sincerity rule: “the speaker commits himself to a belief in the truth of the expressed proposition.” These rules function vertically, in the sense that they establish a connection between language and reality. But in fiction, they are suspended. As a result, the speaker is “able to use words in their usual meaning without undertaking the commitments that are normally required by these meanings” (1975, 326). Novelists are not asserting the facts represented in their novels; they are only pretending to do so, through an overt display of the act of pretense that precludes any deceptive intent. The theories I have discussed so far look at fiction from an external and objective perspective: fiction is made not of true facts but of invented ones; fiction is made not of genuine but of fake assertions. Moreover, they characterize fiction from the point of view of the sender: for Nielsen et al., fiction is something that is used as a rhetorical device; for Searle, writing novels is pretending to do something rather than really doing it. My next theory, Kendall Walton’s Mimesis as Make-Believe (1990), takes instead the point of view of the user. It looks at fiction from the inside, as a particular kind of experience. Rather than speaking of invention, it speaks of fictional truths; rather than speaking of suspension of rules, it speaks of fiction as a game with its own rules. Walton’s theory also departs from others by building on analogy rather than on opposition. “In order to understand paintings, plays, films, and novels, we must first look at dolls, hobbyhorses, toy trucks, and teddy bears. The activities in which representational works of art are embedded and which give them their point are best seen as continuous with children’s games of make-believe” (1990, 11). The analogy characterizes fictions (as well as representations in general, Walton’s broader concern) as “props in a game of makebelieve,” or as “prescriptions to the imagination.” Just as children playing a game of make-believe imagine that stumps are bears, without believing it, the spectators of plays imagine but do not believe that the actors are the characters, the readers of novels imagine rather than believe the events represented in the text, and the spectators of paintings imagine rather than believe that they are facing the objects shown on the canvas. In all these examples, something is taken for something else: the stumps for bears, the actors for characters, the novel for a report of facts, and the painting for the things depicted. In each case, a material object is used to generate “fictional truths,” this is to say, propositions that are true in a fictional world. My last approach is my own conception of fictionality, which combines ideas from Walton and Searle with possible worlds theory, as conceived by the philosopher David Lewis. As we have seen in the introduction, PW theory contrasts a unique actual world (AW) with a plurality of nonactual possible worlds (NAPWs). Since authors are real persons, they are always located in the actual world. The narrators of fiction, by contrast, are located in APWs,

38  •  C hapter 2

but according to Lewis’s indexical conception of actuality (cf. introduction), they tell what are from their point of view true stories about their world. Fictional texts do not tell us “in a world that I [Tolstoy] imagine, Anna Karenina threw herself in front of a train”; they instead tell us “Anna Karenina committed suicide” as a matter of fact. Lewis describes this situation by distinguishing two communicative layers: “Here at our world we have a fiction f, told in an act a of storytelling; at some other world we have an act a’ of telling the truth about known matters of fact; the stories told in a and a’ match word for word, and the words have the same meaning” (1978, 40). The two speech acts are not symmetrical, because they have different reference worlds: the story told at our world, by a real author, concerns an alternative possible world, but the story told in the other world concerns the world where it is told, since it is told as true fact. My own addition to Lewis’s model consists of distinguishing the worldimage presented by the text, which one may call the storyworld in the case of narrative texts, from the world targeted by the author as reference world. All representational texts, whether fictional or factual, project a world-image. Though both fictional and factual representations may present themselves as facts, genuinely factual representations refer to the actual world, while fictional representations do not. The users of factual narratives, after constructing the storyworld, will evaluate this world with respect to their knowledge of the actual world, and they will judge it believable or not.4 In the case of fiction, by contrast, the storyworld is self-referential, and its construction by the user’s mind is an end in itself. The reference world is simply the storyworld. Factual texts can be true or false because they have an external reference world, but it would make no sense to question the facts asserted in a fiction, because the fictional text creates its own world, and it constitutes the only source of infor5 mation about this world. To explain the experience of the user of fiction, let’s go back to Walton’s idea that fiction involves a game of make-believe. This game consists of regarding the fictional world as actual. In order to do so, the users of fiction recenter themselves in the fictional world, and they become in imagination a member of this world. The experience of recentering presupposes a conception of actuality that David Lewis describes as indexical: “Our actual world is 4. Some psychologists claim that readers approach factual texts with an assumption of truth (Currie 2020, 187). If so, it takes really incredible content or a notoriously distrusted author for readers to disbelieve the text. 5. It could be objected that unreliable narrators represent an exception to this claim, but while unreliable narrators may present inacceptable judgments, they must be largely trusted in their factual statements, otherwise the reader could not construct a fictional world. On my conception of unreliability, see note 7 in chapter 3.

F iction  •  39

only one world among others. We call it actual not because it differs in kind from all the rest but because it is the world we inhabit. The inhabitants of other worlds may truly call their own world actual, if they mean by ‘actual’ what we do; for the meaning we give to ‘actual’ is such that it refers at any world i to the world i itself. ‘Actual’ is indexical, like ‘I’ or ‘here,’ or ‘now’: it depends for its reference on the circumstances of utterance, to wit the world where the utterance is located” (1979, 184). Once the fictional world has taken the place of the actually actual world, the characters are no longer viewed as the creations of an author; they become real persons in the imagination. But recentering in a fictional world is not a one-way trip. Users remain aware that they are playing a game, and that they can return at any time to their native reality. While recentering enables them to be emotionally affected by the characters, breaking the game—that is, decentering—enables them to appreciate the author’s performance in creating the characters and the storyworld as a whole. The possibility of traveling inand-out of fictional worlds thus explains both the experience of immersion in fiction and its aesthetic evaluation. Let’s now examine the answers that our five theories give to important theoretical questions. To tease out these answers, I rely on either explicit pronouncements by the authors or on my own interpretation of the implications of their theories.

Is fake news fiction? The naïve theory sees no difference between fiction and fake news, since to be fictional is simply to be false. All the other proposals exclude fake news from fiction: the rhetorical theory, because in fake news deceptive invention is hidden; the pretended speech act theory, because fake news is presented seriously; the make-believe theory, because audiences are asked to believe and not simply to imagine; and the possible worlds theory, because fake news refers to the actual world. From now on I will ignore the naïve theory because it has nothing interesting to say about the next questions.

Does a theory of fiction need a notion of world (fictional, textual, or storyworld)? The rhetorical approach regards fiction as a way to make statements about the real world, and it avoids consequently the notion of a fictional world created by the text. While nonfictional discourse represents the real directly, the

40  •  C hapter 2

rhetorical device of fictionality does so indirectly, so that “fictive and nonfictive discourse represent two options for engaging with the actual world” (Nielsen et al. 2015a, 63). This option explains the rhetorical nature of the proposal: just as I can choose to say something through metaphor or through literal language, I can engage with the actual world through fictional or factual discourse. By doing away with the concept of fictional world, this proposal satisfies the principle of Occam’s razor, since it avoids the postulation of an allegedly superfluous theoretical entity, but for lovers of literature, it reduces the work of the imagination to a didactic function and deprives it of autonomy.6 One wonders, however, how the idea of rhetorical choice works with a novel such as Tolstoy’s Anna Karenina: did the author chose to spin an elaborate tale in order to tell the reader that “happy families are all alike; every unhappy family is unhappy in its own way,” or did he tell the story for its own sake? If this is the case, he had no choice but to make it fiction since it did not happen in reality. The speech act approach, similarly, avoids the notion of fictional world. The author of fiction pretends to make speech acts, but nothing is said about what audiences do with the propositional content of these speech acts. The theory remains therefore uncommitted as to whether audiences construct a world or not out of the propositions that the author pretends to assert. This failure of the speech act approach to address the purpose of the author’s pretense has been a frequent argument against it. In Walton’s make-believe theory, the notions of world and of fictional truth are inseparable. “To speak of a fictional world is, in part, to speak of the class or cluster of fictional truths belonging to it” (1990, 62). The reverse also holds: to speak of fictional truths is to speak of fictional worlds, since there must be a semantic domain for which the propositions imagined by the user are true. This world cannot be the real world; otherwise, there would be no distinction between fictional and factual representation. The notion of world is axiomatic to the PW approach, and it would be circular to try to demonstrate its importance for the theory. Critics of the notion of fictional or storyworld may argue that it is just a metaphor, and that we can do away with it, since there is no other world than the actual world. But sometimes metaphors are the only way to capture an idea. As I explain in the introduction, storyworlds can be conceived of in two ways: as containers for entities that possess a physical mode of existence, and as networks of relations between 6. The authors reject the restriction of the relevance of fiction with respect to reality to a didactic message, but how else can this relevance be conceived? Pleasure taken in the act of imagination inspired by the text does not count as relevant to the real world, because it can be autotelic.

F iction  •  41

these entities. If we conceive storyworlds as containers, then many narratives that take place in the same geographic, historical, and social context will share the same world; for instance, all the novels of Jane Austen take place in the world of the early nineteenth-century English rural upper and middle class. But if we add the requirement that the entities that exist within the container must enter into a network of relations, each of Jane Austen’s novel will project its own world, since the characters of the different novels do not interact with each other. By contrast, a cycle of novels with recurrent characters will project a common world. The ultimate justification for the notion of story­world thus comes from the unity of the representation that the text suggests to the user: if there were no world in which all the characters, objects, and events mentioned in a Jane Austen novel exist, and if there were no network of relations between these entities, the text would be processed as a collection of unrelated propositions. Moreover, by calling the mental construct elicited by text a world, we stress the pretended autonomy of its constituents from the medium of representation. Insofar as the characters, settings, and events of fiction are imagined as real within their world, they are imagined as existing independently from the text, even though the reader knows that they are not.

Does fiction imitate nonfiction? Of the four theories, the speech act approach comes the closest to answering this question in the positive. Searle insists that there is no semantic or syntactic difference between the language of fiction and the language of factual representation. Insofar as fictional speech acts imitate serious speech acts, the difference between fiction and nonfiction resides in an implicit pragmatic operator that relieves the speaker from the responsibility normally entailed by the speech act. However, this operator could very well affect language that, stylistically, does not look at all like nonfiction. A strand of critics inspired by speech act theory have extended Searle’s analysis from the local to the global level by regarding genres as textual speech acts, and by conceiving fictional texts as imitations of the genres of nonfiction. “Novels,” writes Barbara Herrnstein Smith, “have typically been representations of chronicles, journals, letters, memoirs and biographies. Tolstoy writing War and Peace is writing a novel, but his narrator, if there is one, is writing history” (1978, 30). Mary Louise Pratt concurs: novels imitate global speech acts such as biography, autobiography, diaries, and letters. This claim works very well for eighteenth-century novels and for some postmodern ones (for instance, Nabokov’s Pale Fire, which parodies the critical edition of a poem),

42  •  C hapter 2

but, as Pratt observes, with most classic novels (she mentions Jane Austen’s Pride and Prejudice and Thomas Hardy’s The Mayor of Casterbridge), it is hard to identify the nonfictional genre that the text is supposed to imitate. “Does this mean that [these novels] are not really imitation speech acts and that there is no real world correlate for what the fictional speaker is doing?” she asks (1977, 207). Pratt gets around this problem by arguing that the unmarked situation for the novel is to imitate what she calls the “narrative display text,” which is a nonfictional story, typically told in conversation, whose relevance lies in tellability rather than in the transmission of useful information. Today we might say that the distinctive feature of the narrative display text is the attempt to immerse the audience in the storyworld, whether imaginary or not. But the category of narrative display texts is so broad, and its features so diverse, that it does not say much to claim that its fictional imitation constitutes the unmarked case for the novel. The other theories leave open the possibility of a formal and semantic autonomy of fictional texts with respect to nonfiction. If fiction is invention, it is not constrained by the limitations of “natural narrative,” the rather vague concept that forms the standard of comparison of unnatural narratology (Richardson 2016, Alber 2016); if fiction is make-believe, it does not have to look like a text meant to be believed; and if fiction creates, and refers to, a nonactual possible world, it can explore, both stylistically and semantically, a domain far wider than the realm of the actual and believable. It may look at first sight that the formula of David Lewis quoted above, which describes two communicative acts taking place in different worlds but matching word per word, implies an imitation by fiction of the factual narratives of the real world. But this would be a wrong interpretation. What Lewis is saying, rather, is that anything that can be told as true fact in and of a possible world can be told as fiction in the real world, though not of this world. What can be told as true of possible worlds is far wider than what can be told as true of the real world, since there is an infinite variety of possible worlds.

Is fiction a linguistic phenomenon, or does it occur in other media? There are only two media for which the importance of the distinction fiction/ nonfiction (or fiction/fact) is undisputed: the first, of course, is language-based representation. While fiction is in a sense opposed to truth, it presupposes the possibility of truthfulness. Through its ability to express propositions, language has an unrivaled power to claim truth; therefore, it has an unrivaled

F iction  •  43

power to make fictional truth. The other medium that offers both factual and fictional representations is film: there are fiction films that represent performing actors, and there are documentary films that capture real events. Film can be used for fiction because it can also represent real events, and it can represent real events because, like photos, it relies on mechanical capture. The images of film and photos are not only iconic representations, they are also indexical: they capture the patterns of lights emitted by existing objects. As Roland Barthes observed (1981, 76), mechanically obtained pictures give testimony to the existence of what they show.7 Human-made pictures such as paintings and drawings cannot do that: no matter how realistic, no painting can prove the existence of unicorns. But the testimonial value of film can be subverted when it shows actors playing roles rather than people living their lives. This is why the distinction between factual and fictional representation is very clear for language and film, while it is at best problematic for painting. In addition to these two media that can be used for both fact and fiction, there are media that are capable of fiction but unfit for fact: for instance the theater because it relies on actors impersonating characters, and computer games because they involve worlds and objects which are simulated by code, rather than mechanically captured. Moreover, the fact that computer games can be replayed, and generate different worlds where different events take place, disqualifies them from being truthful representations of actual events, since there is only one actual world. While there are media that can be used for fiction but not for fact, there are not, to my knowledge, media that can be used for fact but not for fiction, because if a medium can represent real events, it can also represent simulated or imagined events.8 If fiction is invention, there is no reason why this invention should be limited to language-only media. Fiction film and theater are therefore fully compatible with the rhetorical approach. But film and theater are examples of “generic fiction,” together with literary narrative, and the ambition of the theory is to extend beyond these forms. All the examples of spontaneous, nongeneric fiction provided by Nielsen et al. are verbal interventions, and one wonders whether what the authors regard as nongeneric fiction can occur outside language. Insofar as it is based on speech acts, Searle’s theory presupposes a medium with a language channel. The idea of pretending to perform speech acts applies to actors on a stage, or to film actors as well as to the authors of novels; but it 7. This may no longer be true, now that techniques known as deep fake driven by artificial intelligence can simulate scenes and people in a way that is indistinguishable from actual photos and films. 8. A more detailed version of this argument is to be found in Ryan (2020).

44  •  C hapter 2

does not account for the image component of film nor for the stage setting of drama, and it is inapplicable to silent film or wordless cartoons. By taking children’s games of make-believe as model, Walton’s theory is the most deliberately transmedia of the four. As he writes: “we can learn a lot about novels, paintings, theater and film by pursuing analogies with makebelieve activities like playing house and school, cops and robbers, cowboys and Indians, and fantasies built around dolls, teddy bears and toy trucks” (1990, 4). But the application of Walton’s conception of fiction as “prop in a game of make-believe” leads to a strange asymmetry. In verbal communication, he makes a distinction between fictional texts, which give rise to imagining, and nonfictional ones, which inspire belief or disbelief. But in visual communication, all artifacts are fictional because they all participate in a game of make-believe, by which the spectator forms fictional truths. Viewing a painting of a ship, for instance, the spectator will imagine being in the presence of a ship, rather than viewing a representation; and as she studies the ship, she will form fictional truths such as “this is the mast,” “this is the sail.” I cannot offer here a detailed critique of this view (I have done so in Ryan 2001, 105–10). But I will say this: if all pictorial representations are fiction because they make present objects that are actually absent, then the distinction fiction/ nonfiction becomes irrelevant. And indeed, I think that when we look at a painting from an aesthetic point of view, we do not bother with the question of its fictionality. There have been several attempts beside Walton’s to extend theories of fictionality to media other than literature, but when they address painting (Currie 1990, Schaeffer 1999), these attempts either link visual fictionality to the nonexistence of the depicted object (“this picture is fictional because it represents a unicorn”), which amounts to reverting to what I have called the naïve definition, or they make visual fictionality dependent on the verbal kind, by regarding it as the illustration of fictional texts. While it is desirable for a theory of fictionality to extend beyond language, this does not mean that the distinction fiction/nonfiction is valid for all media. The PW approach, in its initial formulation, is more focused on verbal storytelling than Walton’s theory is. If we adhere closely to Lewis’s distinction between two acts of storytelling, one taking place at our world, the other in an APW, then it has a strong linguistic and narrative bias. To extend it to theater and film, it should be rewritten as “Here at our world we watch actors impersonating characters; at another world, we are witnesses to people living their life.” I haven’t found a way to define fiction through a single formula that covers both the telling and the showing forms, yet the idea of imaginative recentering into a world where the represented events actually happen remains valid for both narrated and enacted forms of fiction.

F iction  •  45

Are counterfactuals and thought experiments fiction? Counterfactuals are “if .  .  . then” statements by which people imagine what would have happened if some verified events had not taken place. They are particularly popular with sports fans (“If not for that bad referee call, our team would have won the game”) as well as with historians, who like to debate questions such as what would have happened if John F. Kennedy had not been assassinated (Kunz 1997). Through such reasoning, sports fans and historians locate turning points and propose explanations. As for thought experiments, they are imaginary scenarios that are invoked in philosophical discourse or in science popularization in order to explain certain concepts, or to criticize certain interpretations. A classical example is the scenario of Schrödinger’s cat, used by Schrödinger in a paper about the current state in quantum mechanics to demonstrate the absurdity of the idea of a superposition of states. The imaginary nature of the scenario is acknowledged in the introductory phrase “One can even set up quite ridiculous cases” (Schrödinger [1935] 1980, 328). Submitting a live cat to radioactive decay is not, one hopes, something that real scientists would do! For the rhetorical/invention approach, counterfactuals and thought experiments represent local fictionality in globally nonfictional discourse. They are a way to invoke the nonfactual in order to make points about the real world. In so doing they fully represent the notion of fictionality, since according to the theory all fiction is ultimately about reality. The thought experiment of Schrödinger’s cat clearly involves invention, but with a counterfactual statement such as the comment by the sports fans, one wonders what is “inventive” in denying an actual fact (the referee making a bad call) and following the consequences. Searle’s theory would not regard counterfactuals as fiction, since they are serious speech acts and do not involve pretense. Walton does not address the issue of counterfactuals and thought experiments, but it is a safe bet that he would not regard them as “props in a game of make-believe,” since their point is to make believable statements: what caused our team’s defeat in the case of the sports fan, how absurd is the idea of a cat being both dead and alive in the case of Schrödinger’s cat. Similarly, for the PW approach, counterfactuals cannot be fiction, because they do not deploy a possible world for its own sake but instead invoke the nonfactual as a way to make a point about reality. The reference world of counterfactuals is therefore the real world. Counterfactuals differ from fiction not only through their reference world but also through the syntactic device if . . . then, which makes obvious their lack of reality and prevents them from

46  •  C hapter 2

being told as true facts.9 But a distinction should be made between the use of counterfactuality by historians and novels of counterfactual history, which present the counterfactual as fictionally true. (A classic example is Philip Roth’s novel The Plot Against America.) Similarly, thought experiments differ from standard fiction through the use of semantic devices such as “imagine that” or “consider this ridiculous case,” which function as command to the imagination, and expose them as invention. It could be argued that through paratexts such as “novel” or “short story,” literary fiction also exposes its nonfactual status, but there is an important difference: thought experiments are inserts within nonfictional discourse, and the marker of unreality frames them from their surroundings, whereas the paratext “a novel” is external to the fictional text. The unreality is meant to be bracketed out once the reader enters the fictional world. Not so with thought experiments: their value resides in how well they support the arguments made in the surrounding text. Moreover, in their subordination to a thesis in need of demonstration, they are usually too sketchy to project a world, much less a storyworld that stimulates the imagination.10

Can fiction make truth-claiming statements about real-world entities? This problem is particularly acute in historical fiction, such as Tolstoy’s War and Peace, which includes characters who really existed, such as Napoleon or General Kutuzov, and says things about them that are historically verifiable. According to the rhetorical approach, “Global fictions can contain passages of nonfictionality, and global nonfiction can contain passages of fictionality” (Nielsen et al. 2015a, 67). Historical novels can consequently be regarded as global fictions that contain local truths, and a work like War and Peace is a patchwork of invention and verified facts. Searle’s pretended speech act approach also regards fictions that refer to real-world entities as patchworks—in this case patchworks of pretended and serious speech acts: Most fictional stories contain nonfictional elements: along with the pretended references to Sherlock Holmes and Watson, there are in Sherlock 9. Though counterfactuals and fiction are different discourse modes, David Lewis’s 1978 account of truth in fiction is derived from his 1973 account of the truth conditions for counterfactuals. 10. See Currie (2020, 135–49) on the differences between literary fiction and the thought experiments of philosophy and science.

F iction  •  47

Holmes real references to London and Baker Street and Paddington Station; again, in War and Peace, the story of Pierre and Natasha is a fictional story about fictional characters, but the Russia of War and Peace is the real Russia, and the war against Napoleon is the real war against the real Napoleon. 11 (Searle 1975, 330)

The test for telling the fictional from the nonfictional is what can be potentially regarded as a mistake and what cannot: “If Sherlock Holmes and Watson go from Baker Street to Paddington Station by a route which is geographically impossible, we will know that Conan Doyle blundered, though he has not blundered if there never was a veteran of the Afghan campaign answering to the description of John Watson, M.D.” (Searle 1975, 331). But did Conan Doyle blunder when he gave Sherlock Holmes the address of 221B Baker Street, an address which does not exist? And, more generally, are readers of the Sherlock Holmes stories on the lookout for what is true and what is not, for blunders and accurate statements? The world-based theories argue that such approaches do not account for the experience of the reader. To the implicit segregationism of worldless approaches, they oppose an integrationist view (terms borrowed from Pavel 1986a). Fictional worlds are normally homogeneous entities,12 and readers accept the statements that describe them as globally true in make-believe. When historical characters, such as Napoleon, appear in fiction, their properties can be partly verified, partly invented, but this does not affect the unity of the character. Some critical schools (historicism, positivism) may be interested in sorting out the true and the false, but most of the time readers are unable to do so, and they do not really care, because what matters to them is fictional truths. When authors use the names of characters or locations that exist in the real world, they are referring not to the real entities but to their counterparts in another world, and they are not committed to telling real-world truths. According to PW theory, possible worlds may be located at variable distances from the real world, and their inventory of individual characters may overlap to some extent with the inventory of the real world. The Napoleon of War and Peace is not, technically, the Napoleon of the real world, but the real Napo 11. A real reference would mean that Conan Doyle himself refers to London. But in this argument Searle seems to forget that it is Dr. Watson, a fictional narrator, who does the reference. Insofar as Conan Doyle pretends to be Dr. Watson, this is a case of fictional (=pretended), not real reference. The argument is more credible for War and Peace if one adheres to a narrator-optional theory (ON). In the absence of an individuated narrator, it could be said that Tolstoy himself refers to Napoleon, though I am skeptical of this interpretation since I do not adhere to ON theory. (See chapter 3.) 12. Except in the case of fictional worlds consisting of ontologically distinct domains across which characters can travel, such as the many-worlds ontologies described in chapter 7.

48  •  C hapter 2

leon is linked to the Napoleons of other possible worlds through counterpart relations. Some of these other Napoleons are fairly close to the real one; others have vastly different properties, such as winning the battle of Waterloo or escaping from Saint Helena. From an external perspective these other Napoleons are partly real and partly invented, but from a world-internal perspective they are fully real, and everything the text says about them is true in the storyworld. Moreover, thanks to what I have called the principle of minimal departure (Ryan 1991), readers are entitled to attribute to fictional Napoleons everything they know about the real one, except for the information that is contradicted by the text. This does not mean that world-based approaches exclude the possibility of combining facts and fiction in the same work. But for such combinations to occur, the two components must be clearly framed from each other. For instance, the novel Lincoln in the Bardo, by George Saunders (2017), combines clearly identified quotations about Abraham Lincoln taken from historical works with a fantastic narrative about people buried in the same cemetery as Lincoln’s young son who exist in a state of being neither dead nor alive, and who witness with amazement and envy Lincoln’s visits to his son’s grave. The novel superposes two worlds—a historical world, where the quotations exist but the ghosts do not, and a fictional world, where both the quotations exist and the dialogues of the ghosts truly happened. But because of the explicit framing, this situation is quite different from regarding regular historical novels as patchworks of facts and invention, or of serious and pretended speech acts. The distinction between segregationist and integrationist approaches has important consequences for two phenomena. The segregationist stance is better able to explain how people can learn about the world from reading fiction. Readers of fact-based fiction assume that authors have done their research and/or rely on their life experience, so that the information reported in the text does not entirely come from the imagination. It is always risky to extract knowledge from fiction because the true and the invented are not marked differently (as Lewis suggests, they are both told as true), yet it is their desire to learn about history, or about foreign cultures, that drives many readers toward certain fact-based novels.13 On the other hand, the integrationist approach is better at explaining immersion because it offers to the reader a homogenous world in which people can relocate themselves in imagination. By contrast, a 13. On the question of learning facts from fiction, see Currie (2020). Currie believes that a text like War and Peace can teach us some individual facts, such as “the Russian aristocracy spoke French,” but he is skeptical about fiction teaching more general and therefore meaningful truths, such as truths about “the human heart.”

F iction  •  49

segregationist approach that distinguishes between invented and true statements requires critical distantiation. It is interesting to note in this respect that Richard Walsh, one of the proponents of the rhetorical approach, explicitly rejects immersion as a valid type of textual experience: immersivity, for Walsh, is a property of certain media, such as virtual reality or maybe film, not a property of certain representations (2017, 476).

Is the border between fact and fiction gradual (scalar) or binary? In classical logic, the truth of a proposition is a binary matter: it is either true or false that Napoleon died on Saint Helena. Yet there are predicates that involve a scalar conception of truth: for instance, whether Napoleon was short or not is a matter of degree. On the textual level, similarly, truth comes in different degrees, and users adjust their expectation of factuality according to the genre. Among nonfictional texts, the highest degree of factuality comes in genres such as scientific writing, history, or court testimonies; the lowest degree, in autobiography and in conversational narratives. Autobiography contrasts with biography and history in that authors can make assertions about themselves without documenting them, since they are the only ones to know. The unreliability of memory makes it often impossible to tell what is true and what is not, especially when authors dwell on deeply private matters. Conversational narratives such as gossip or narratives of personal experience are supposed to be true in their broad lines, but they must also entertain the audience. These conflicting goals often lead storytellers to play loose with the truth. For instance, dialogues enliven the performance, but audiences do not expect the storyteller to remember conversations precisely: it is good enough to report what people could have said. Yet another type of text that questions rigid distinctions between fact and fiction is the genre (or genres) variably known as true fiction, nonfiction novel, faction, or creative nonfiction. One of the best-known examples of this genre is Truman Capote’s In Cold Blood, which reports a horrifying murder in Kansas through novelistic techniques, such as extensive dialogues, reports of private thoughts, and vivid descriptions. These techniques lead to effects that are typical of novels but not of history or traditional journalism, namely the effects listed in the promotional text on the back of the book: “mesmerizing suspense” and “astonishing empathy.” Weak factuality is easily accepted by readers because it is compensated by an increase in tellability and immersivity, but it presents a dilemma for theories of fictionality: is the distinction between fictional and factual narratives gradual, or binary?

50  •  C hapter 2

The rhetorical approach is ambiguous about whether to draw a rigid border between fictionality and factuality or not. Here is a statement that suggests a gradual theory: “From our perspective it is wiser to talk about degrees of fictionality rather than the distinction of fiction” (Nielsen et al. 2015a, 67). And here is one that supports a rigid distinction: “The fictional status of communication is not a matter of degree but rather the consequence of the communicator’s intent to use invention (or not)” (Nielsen et al. 2015b, 104). A conceivable way to justify these contradictory claims could be by saying that fictionality is a binary operator on the global level, signaled to the user in many ways: by the communicative situation (such as a political rally, classroom teaching, or tall-tale contest), by the platform of transmission (such as newspaper, scientific journal, or science fiction magazine), or by a paratextual indicator such as the label novel. But on the local level, texts can be more or less fictional, since they can consist partly of invention, partly of true facts. Thus, War and Peace would be a global fiction because of the generic label novel, and so would Lord of the Rings, but Lord of the Rings would be more fictional because it presents a higher proportion of invention. This makes fictionality both binary and gradual. But the fictionality of the local level lacks the crucial property of being openly signaled: authors do not mark invented elements differently from verified facts. Readers would consequently be on their own to decide what is invented and therefore fictional and what is not, if such a sorting out were indeed necessary to the proper comprehension of the text. Non-­segregationist theories would argue that it is not: both verified and invented facts contribute to the construction of the storyworld, and there is no such thing as a “local” nonfictionality in globally fictional texts unless properly framed. Searle’s theory allows serious speech acts about real-world entities such as London in a globally fictional text. But what does the theory do with the unverifiable statements of a globally fact-based text like In Cold Blood? Is the author asserting them, like the verifiable facts of the text, or just pretending to do so? In a text like War and Peace, the serious speech acts concern realworld referents, and they are easily distinguished from the pretended speech acts, which concern imaginary objects, but in a text like In Cold Blood, all the assertions concern existing entities, so there is no formal way to decide what is pretended and what is not. Searle’s model does not know what to do with borderline texts such as weak factuality and true fiction. Walton’s theory says that texts can be either meant to be believed or used as props in a game of make-believe. These attitudes define, respectively, factual and fictional works. But from the point of view of the reader, the distinction between fiction and nonfiction is not absolute: “Some histories are written in

F iction  •  51

such a vivid, novelistic style that they almost inevitably induce the reader to imagine what is said, regardless of whether or not he believes it. (Indeed this may be true of Prescott’s History of the Conquest of Peru.)” (Walton 1990, 71). By “imagining” what is said, Walton means something like transporting oneself in imagination to the scene of the events, and becoming in make-believe a witness to their unfolding. If a text satisfies both our need for acquiring true information and our need for aesthetic gratification, which resides in the pleasure taken in the play of the imagination, it should be considered both fictional and factual. But isn’t this a logical contradiction? While Walton would say that a nonartistic historiographical text would elicit belief (or disbelief) but no imagining, I claim that all mimetic texts ask the user to imagine a world; in addition to this operation, users of factual texts decide whether to believe them, or parts of them, or not.14 In my model, the existence or nonexistence of an external reference world decides whether the text is fictional or factual; this is a binary criterion. Conversational narrative, even though not entirely believable, clearly takes the real world as reference world. By contrast, fictional texts create their own world rather than referring to our actual world. According to PW theory, possible worlds can stand at variable distances from the actual world: the worlds of fantasy are very remote, while the worlds of realistic and historical fictions are fairly close. A text like In Cold Blood could represent the most extreme forms of proximity between a fictional world and the actual world. But this solution overlooks the interest of the reader for the text’s factual content. In Cold Blood would lose much of its appeal if it did not report a true crime. While the author is free to use his imagination to enliven the narration, he remains under the obligation to respect the known facts because the reader’s interest lies as much in the relation of the storyworld to the actual world as in the storyworld itself. This situation can be accounted for by giving the storyworld a double reference: on one hand, it provides information about the actual world; on the other, it is self-referential. But like Walton’s claim of a double status with respect to fictionality, this may look like an ad hoc solution. The idea of a fuzzy boundary between fact and fiction, which implies a nonbinary conception of the two types, provides an easy and intuitive solution to the categorization of texts like In Cold Blood: it has become standard practice for commentators (especially for those who are not particularly well 14. This thesis is also defended by Derek Matravers. Against Walton’s distinction between imagining and believing, Matravers claims that from a cognitive point of view, “imagination is not the counterpart of belief; indeed, in many cases there is no distinction between imagination and occurrent belief ” (2014, 3). For Matravers, both fictional and factual texts give the reader something to imagine, and the distinction between the two types lies elsewhere.

52  •  C hapter 2

versed in theories of fiction) to write that a given text “blurs the borderline between fact and fiction.” In a nonbinary model (figure 1a), there is a continuum of texts based on the strength of their claim to truth. This continuum runs from science and history, to conversational narratives and autobiography, to true fiction, to realistic fiction, and, finally, to fantastic texts. Insofar as it regards realistic or historical novels as less fictional than fantastic tales, the gradual model classifies texts according to their relation to reality, but it does not take into consideration the author’s intent, nor the user’s awareness of this intent. Therefore, it cannot distinguish fiction from lie and error. Fictionality is not a function of the degree of truth of a text with respect to reality; it is a matter of deliberate framing. A text presented as fact can also, at least in principle, be presented as fiction, as we can see from the case of A Million Little Pieces, by James Frey (2005). This text, which chronicles the narrator’s drug addiction and recovery, was originally published as a memoir and selected by the Oprah Winfrey book club for the nonfiction category.15 But a scandal erupted when inaccuracies and fabrications were discovered. In other words, the book did not fulfill the truth requirements of a memoir. Nowadays it is published as a novel, and though the text is exactly the same, the author can no longer be blamed for telling lies. On the other hand, My Struggle, by Karl Ove Knausgaard (2013–19), could have been presented as a memoir because it consists mainly of autobiographical material, but the author chose to call it a novel. These examples demonstrate that the divide between factual and fictional narrative is far more rigid than the continuum model suggests, and that it does not depend on the truth of the text. A binary approach is shown in figure 1b. By proposing their work as either fact or fiction, authors make an ethical commitment to the truth, or, on the contrary, they refuse to do so. It is therefore entirely possible for a basically true text to be presented as fiction. On the left side of the border, texts propose an image of the real world; on the right side, they create a world for its own sake. The left side of the border presents a continuum from strong to weak factuality. But on the right side, there is no continuum from weak to strong fictionality. Fictional worlds can be more or less distant from the real world, which means more or less materially possible; but a realistic novel is just as fictional as a fantastic tale. While there are degrees of factuality, there are no degrees of fictionality. 15. Actually, Frey originally conceived his text as a novel. As Helena de Bres writes, “There are enough fake occurrences and major exaggerations in Frey’s A Million Little Pieces (more impressively, he converted a few hours at a police station into a three-month prison stay) that he originally tried to sell the book as a novel, before rebranding it as a memoir. Random House accepted the memoir with zero fact-checking after having rejected the same manuscript when submitted as fiction” (2021, 95).

F iction  •  53

FIGURE 1. Gradual and binary theories of fiction

But what are we supposed to do with texts that challenge the borderline by combining features of both factuality and fictionality, such as In Cold Blood, or with literary experiments that create collages of openly fictional and factual fragments, such as Lincoln in the Bardo? One way of accounting for these texts would be to postulate some kind of no-man’s land between fact and fiction, a no-man’s land where texts are neither fact nor fiction (figure 1c). This solution is inadequate because it gives readers no reason to be interested in the text: it is not fact—so it will not improve their knowledge of the real world, and it’s not fiction, so imagining the storyworld will not be a self-rewarding activity. The alternative to the no-man’s land is to make the territories of fiction and fact overlap (figure 1d), so that hybrid texts can find a home, without giving up the distinction between fact and fiction. Some texts are created to elicit belief, others to prompt self-rewarding imagining, but these effects are not mutually exclusive. A text that clearly benefits from the postulation of an overlap of fact and fiction is Art Spiegelman’s graphic narrative Maus ([1973] 1986), which recounts the life of Spiegelman’s father, an Auschwitz survivor. A case can be made for Maus as a historical work: Spiegelman recorded his conversations with his father, and he stayed as close as he could to his father’s testimony. For many people, the testimonial value of the narrative is a major

54  •  C hapter 2

source of its appeal. But a case can also be made for the fictionality of Maus. The text presents a world inhabited not by humans but by cats and mice and dogs and frogs. These characters behave like humans, but they still look like animals. So how does the reader imagine the storyworld of Maus? Is it a world inhabited by anthropomorphic animals, like the worlds of La Fontaine’s or Aesop’s fables? Or is it a world inhabited by Nazis who pursue Jews like cats hunt mice? In this case one could say that in the storyworld of Maus there are no cats or mice or other animals (except for the attack dogs kept by the Auschwitz guards); there are only Nazis and Jews and other identities. The visual representations of Nazis as cats and Jews as mice would be a visual metaphor: when somebody says of my friend Bill that he is a donkey, I do not imagine Bill with long gray ears; I assume that the speaker means that he is stupid. But the interpretation of the cats and mice as visual metaphors ignores the graphic nature of the medium. Visual media give less freedom to the imagination than language-based ones because unlike language, they consist of sensory data that force the mind to visualize things the way they are represented in the text. We cannot make the cats and mice of Maus disappear from the imagination by saying that they are just metaphors, or allegories, because they are the main source of the text’s artistic innovation. This ambiguity explains why nobody has an easy time classifying Maus as either fact or fiction.16 As Marianne Hirsch notes, “The Pulitzer prize committee invented a special category for Maus, suggesting the impossibility of categorizing it as either fiction or nonfiction” (1997, 274). Nancy Pedri (2013) writes that such texts are neither fact nor fiction (figure 1c), but as I have already suggested, this verdict gives readers no reason to be interested in Maus. The overlap of diagram 1d accounts for the fact that some people view the text as factual, oth17 ers as fictional, and still others as both. 16. For a defense of Maus as factual narrative, see Baroni (2021). Baroni defends the thesis that all media are equally capable of factuality and fictionality. For a diverging opinion, see Ryan (2020). 17. Where on this diagram do I propose to locate autofiction? This term has recently become popular for designating contemporary experiments with life-writing that blend truth and invention, but it covers a variety of cases, from memoirs resorting to novelistic techniques (Tara Westover’s Educated [2018], which tells about the author’s growing up in a Mormon survivalist family in Idaho and her struggle to get an education) to heavily autobiographical novels (Karl Ove Knausgaard’s My Struggle) to the creation of counterparts of the author who undergo totally imaginary events (the murder of a character named Michel Houellebecq in Houellebecq’s La Carte et le territoire [2010]). The categorization of life-writing depends on where the interest of the reader is located. In the case of the autobiography of a celebrity, the reader wants facts more than literature, and these texts represent the low-factuality end of the nonfiction domain. People read Educated primarily because it tells a true story, but also because it reads like a novel; I place it therefore in the zone of overlap. Other examples of the overlap are

F iction  •  55

This solution is admittedly not very elegant, but it is the only option left if, as I hope to have shown, we need to dismiss the other possibilities. No respectable theory of fictionality has a really convincing solution for cases like Maus or In Cold Blood, because in order to do so the theory would have to be at the same time gradual and binary: binary in order to describe fiction as the product of a contract between sender and receiver, but gradual to account for combinations of documented facts and made-up elements. This failure explains why many readers will be satisfied with regarding certain texts as indeterminate with respect to fictionality or factuality. It is indeed far less important to categorize In Cold Blood or Maus as either fiction or history than to do so with War and Peace and Napoleon’s Invasion of Russia, by George Nafziger (1998), because these last two texts are safely situated outside the controversial zone, and their categorization has clear consequences for the reader’s knowledge and beliefs.

Conclusion Each of the five theories I have discussed projects a different light on the phenomenon of fictionality. But each of them also presents an Achilles heel, and this leads to unsatisfactory answers to some of the questions we may want to ask of them. The Achilles heel of the naïve theory lies in its inability to exclude fake news and lies from the domain of fictionality. For the rhetorical theory, vulnerability resides in its failure to tell the difference between utilitarian and autonomous uses of invention, this is to say, between invention used to say something about our world, and invention presented for its own sake. By insisting on real-world relevance, the model works clearly better for utilitarian types of fiction, such as parables and thought experiments, than for pleasureoriented texts such as novels and films. Moreover, as the “Romneysia” example shows, the rhetorical approach confuses fictionality with verbal creativity, and ironically—given its moniker rhetorical—cannot distinguish fictionality from rhetorical figures such as metaphor. The pretense theory, as Searle envisions it, suffers from immersive deficiency, due to the lack of world concept and to the segmentation of texts into serious and pretended speech acts—a deficiency it shares with the rhetorical theory. Moreover, by presenting fiction as Jean-Jacques Rousseau’s Confessions and Elie Wiesel’s Night. By choosing to label his work a novel, Knausgaard builds a world that stands very close to the actual world but is nevertheless fictional; readers are interested in the character Knausgaard, but they don’t really care about whether the text offers an accurate image of his real-world counterpart or not. And, finally, Houellebecq’s La Carte et le territoire falls squarely into the fictional domain.

56  •  C hapter 2

pretended assertion, Searle’s definition is too limited to language-based representation. Walton’s conception of fiction as “prop in a game of make-believe” is built on a questionable opposition between believing and imagining. How can we decide whether to believe a representation or not if we do not begin by mentally picturing it, which means imagining? Moreover, the extension of the concept of make-believe to the visual domain makes the dubious claim that all images are fictional. The Achilles heel of my own application of PW theory lies in the notion of world. If we conceive possible worlds as coherent totalities, then when a text asserts contradictory propositions, it does not project a world. But there are texts that deliberately cultivate contradiction: nonsense rhymes, and some postmodern novels. These texts are not immersive, but they are certainly fictions. Should one then conclude that a fictional world does not have to be logically consistent, in contrast to a possible world (this would be Walton’s answer), or can there be worldless fictions? I return to this problem in chapter 8. In contrast to the rhetorical theory, the PW approach does better with entertainment-oriented fiction than with the utilitarian types, since it regards fictional-world creation as autotelic. But by lifting the requirement that fictional worlds be contemplated for their own sake, the PW approach can be made to account for the utilitarian types of fiction. Compensating for these shortcomings are the strengths of the theories. The rhetorical theory opens up the issue of fictionality to a greater debate, summarized by this quote: “The ability to invent, imagine, and communicate without claiming to refer to the actual is a fundamental cognitive skill, one crucial to humans’ interactions with their world and their fellow beings in that world” (Nielsen et al. 2015a, 63). Through its distinction between fictional and nonfictional statements in a fictional text, the proposal also explains why at least some people may read fiction to learn precise facts about the real world. Searle’s pretended speech act theory, by relieving the author from the responsibility of asserting the facts represented in the story, provides the best account of the ethical implications of the decision to present a text as fiction. By labeling their creations novel rather than memoir, authors of texts based on personal recollections renounce the strictest level of responsibility for truth and give themselves greater freedom of expression. Walton’s theory frees fictionality from language, and places it within the larger context of the importance of play in human life. The PW-inspired theory gives a phenomenological account of the experience of immersion as imaginative relocation within a world. It also supports the phenomenon of transfictionality and transmedial storytelling (cf. chapter 10) by allowing multiple texts to refer to the same world, something the worldless theories, for obvious reasons, would have a hard time explaining.

F iction  •  57

For a theory of fictionality to stand a chance of being valid, it must fulfill two conditions: first, it must distinguish fiction not only from truth but also from failures of truth due to error or deception. Second, it must account for the prototypical forms of fictionality, which is narrative fiction, as found in literature, some forms of oral storytelling, drama, and film. In their attempt to do so, theories may catch a wider range of phenomena than prototypical fiction. These extensions can be a plus because they situate prototypical fiction within a larger set of human behaviors, and in so doing make important claims about the functioning of the human mind. We could not enjoy narrative fiction if our brains were not wired to consider the nonfactual, and we could not create fictional worlds and enjoy them for their own sake if we did not do it as children in our games of make-believe. But an ideal theory should be able to capture both differences and similarities between the prototype and the extended field. It is my belief that theories that rely on the notion of world are better equipped to capture the experience of literary, filmic, dramatic, and game fiction than worldless theories because the notion of world gives a home to the imagination, and a purpose to its activity: the purpose of exploring this home.

CHAPTER 3

Narrator Decomposing a Theoretical Primitive

When Barthes and Foucault opened up the notion of author to examine its inner organs, the operation was generally considered successful (to judge by the critical fortune of their ideas), but the patient died. The narrator may seem to be protected from this fatal deconstruction by an obvious logical argument: if narrative is a type of message produced through an act of narration, this act must be performed by an agent; and what should this agent be called if not a narrator? Yet, while the existence of a narrator is uncontested for first-­person narrative fiction,1 it has been challenged for certain kinds of third-­person narratives, first in the 1980s by Ann Banfield’s and S.-Y. Kuroda’s non-­narrator theory, then in Richard Walsh’s Rhetoric of Fictionality (2007), and more recently in Sylvie Patron’s optional-narrator theory (2016, 2020). These proposals have the merit of inviting narratologists to rethink the notion of narrator, which is too often taken as necessary, given, monolithic, and self-evident. 1. The terms first-­person and third-­person narration have been criticized by Genette (1980, 243) for a lack of symmetry: first person refers to the narrator, third person to characters. But I will keep using the terms, first because they are well established in narratology and, second, because if one regards the “person” as referring to the characters—as Genette recognizes— rather than to the narrator, these designations are perfectly appropriate: in first-­person narration the main character is referred to as I, and in third-­person as he or she. This interpretation also justifies the use of second-person (you), first-­person plural (we), and third-­person plural (they) narration. However, it leaves a gap when the narrator speaks as I but the main character is somebody else, as in Thomas Mann’s Doctor Faustus. 58

N arrator  •  59

This appearance of simplicity is strengthened by the tautological character of most definitions: Gerald Prince’s Dictionary of Narratology defines the narrator as “the one who narrates, as inscribed in the text” (2003, 66), while Webster’s dictionary does not even bother with a definition: narrator is simply the agentive noun derived from the verb to narrate by adding the suffix -or. In the first part of this chapter, I review the arguments for and against the narrator-optional theory of fiction. In the second part, I propose to demonstrate that the notion of narrator is not the theoretical primitive for which it has been taken so far. “Narrating” a story is a complex activity which can, and should be, analyzed into distinct functions. “Narratorhood” is therefore a matter of degree: the presence, visibility, and psychic density of the narrator depends on how many of these functions are fulfilled by the storytelling agent.

Narrators: Mandatory or Optional? Narrator-based theories of narrative—theories that postulate a narrator for all kinds of narration—present the advantage of allowing the distinction between fictional and nonfictional narratives through a simple formula proposed by Genette (1997, 73): in factual narrative, A = N; in fiction, A  N. (A and N stand for author and narrator.) Moreover, the formula A  N brings under a common denominator two forms of fictional narration, those in which the narrator is an individuated character and narrates in the first person, and those in which the story is told in the third person and does not make a narrator visible. This model has recently come under attack by theorists who deny the presence of a narratorial figure in third-­person narration, and who therefore regard this figure as an optional feature of narrative fiction. In the following discussion, I refer to the two kinds of theories as MN (mandatory narrator) and ON (optional narrator). ON theories rest on arguments like these (I gather them from various authors; when no name of proponent is mentioned, this means that the argument is openly or tacitly endorsed by most ON proponents rather than author-specific): 1. Third-­person narration does not create narrators who exist as individuals within the storyworld. Therefore: 2. Readers do not imagine narrators as persons in third-­person narration. 3. MN theories lead to the postulation of a “monster”: the omniscient narrator (Patron 2020, 4).

60  •  C hapter 3

4. The postulation of a narrator figure in third-­person narration means that this narrator is responsible for “selecting, arranging, and communicating the events of the story” (Boyd 2017, 285), and it thus prevents the appreciation of the artistic performance of the author. 5. Third-­person narration does not correspond to any genre of nonfictional narration. 6. The postulation of a narrator leads to the false question “how does he know?” In third-­person narration, the discourse establishes the facts and does not need to be validated. 7. Many third-­person narratives could not be uttered by a narrator, because they include “unspeakable sentences,” such as sentences that represent the thought or speech of characters in the free indirect mode (Banfield 1982). These sentences, though representing a character’s point of view, are uttered in the third person, presupposing that the character is not the narrator. They cannot therefore represent the subjectivity of a narrator.

The basic assumption of ON theories is that narrators must be embodied humans (or anthropomorphic creatures), and that if narrative discourse cannot be attributed to such a being, it has no narrator. If we abandon this assumption, the arguments listed above can be refuted as follows (numbers refer to the list above): 1. The term narrator denotes the agent of an act of narration. It does not imply characterhood. “Narrator” can be a textual function rather than a person, an abstract theoretical entity rather than an individuated member of the FW. 2. If narrator is an abstract theoretical entity rather than an imaginary person, there is no need to imagine “it” as an individual. 3. The term omniscient narrator refers not literally to an individual with superhuman abilities but (metaphorically) to a mode of narration that reports as fact what human observers can only guess: the thoughts of other people. (Would omniscient narration be more acceptable to ON advocates?) 4. If the presence of a narrator prevents the appreciation of the author’s story­ telling skills by attributing the organization of the material to this narrator, then the same would hold for first-­person narration, such as Camus’s The Stranger. If readers are able to appreciate Camus’s narrative art in The Stranger, there is no reason why the postulation of a narratorial discourse for a third-­person narrative would prevent them from doing so.

N arrator  •  61

5. As I argue in the introduction, this is also the case for many first-­person narratives: for instance, the Benjy section in Faulkner’s The Sound and the Fury, or the narrator of Beckett’s The Unnameable. 6. If the third-­person narrator is a function, an abstract theoretical entity, this problem does not arise. The author makes up the facts, but the narrator reports them as if they existed independently of the discourse. Or to put this differently: the third-­person narrator tells as fact what the author imagines, and it does not need to justify its knowledge. 7. Again, this presupposes that the narrator is an individuated person. If the narrator is conceived as an abstract theoretical entity, it is not limited by linguistic conventions nor by pragmatic conditions concerning what people can believably say. Moreover, the linguistic argument is highly language-dependent and would not hold for a language that does not present the characteristics of free indirect discourse.

A major problem for ON theories is that there must be a source for narrative discourse because this discourse is made of language, and language consists of symbolic signs which, in contrast to indexical and iconic signs, cannot exist by themselves in the world. An index, such as smoke meaning fire, is not intentionally produced but is rather part of a causal process that unfolds independently of human agency. As for iconic signs, while the vast majority represent human attempts at communication, this is not necessarily the case: a cloud occurring in nature could suggest an animal to the observer. But the signs of language must be produced by an agent, whether this agent is a fully fleshed character in a narrative or an abstract, theoretical construct. Language-based narrative differs in this respect from the visual kind, such as film, drama, or graphic novels. Visual narratives give something to apprehend to the senses, and spectators can imagine that they are directly witnessing events, independently of any mediation. It is therefore not necessary—and in my view clearly wrong—to attribute visual narrative to a narratorial figure. On the other hand, if language cannot exist by itself, immediacy cannot occur in language-based narratives. These narratives may occasionally convey a sense of direct presence, described by critics as “showing” as opposed to “telling” (see chapter 6), but showing is an illusion: language-based narratives are always told, and telling needs an agent. If we assume that language-based fictional narrative comes in two variants, first-­person homodiegetic and third-­ person heterodiegetic, ON theories associate third-­person heterodiegetic with visual narrative, as opposed to the first-­person kind, while MN theories associate third-­person narrative with the first-­person kind, and opposes them both to the narratorless visual kind, as shown in figure 2.

62  •  C hapter 3

FIGURE 2. How narrator theories group together fictional forms

For ON theories to be viable, then, an explanation must be found for how language can exist without a source. S.-Y. Kuroda, an eminent linguist who was one of the first proponents of ON theories, argued that language can be freed from the communicative function that links a sender to a receiver, or a narrator to a narratee. But postulating noncommunicative instances of language would be an ad hoc solution to the problem of third-­person narration if it occurred only in fiction. Kuroda therefore proposes other examples: soliloquy (where the receiver is the sender), magical incantations (directed at supernatural creatures), or, controversially, legal decrees (Patron 2020, 10). These examples may lack a distinct human receiver, but they do not lack a sender, or the utterance would not exist. They may therefore not be communicative in an ordinary sense, but they are certainly expressive of a sender’s intent. If we transfer this reasoning to third-­person narrative, we can say that these narratives may not have a narratee but they must have a narrator. Another problem with the postulation of noncommunicative uses of language lies in how sentences can convey meaning if they are not produced by a sender. Kuroda answers this problem by postulating an “objective” function of language, though which sentences can mean without participating in communicative acts. This proposal clashes with H. P. Grice’s account of nonnatural (that is, symbolic) meaning: for A (a sender) to mean something by x (a material sign, or signifier in Saussure’s terminology), “A must intend to induce by x a belief in an audience, and he must also intend his utterance to be recognized so as intended. But these intentions are not independent; the recognition is intended by A to play its part in inducing the belief ” (1971, 441). If one accepts Grice’s analysis, language cannot convey meaning without expressing an intent.

N arrator  •  63

ON has a very simple answer to the problem of the source of narrative discourse in third-­person narrative: this source is the author. No other source is needed. But the author is also the literal, or material, source of first-­person narrative, so in this case we have a duplication of sources, since the fictional narrator functions as producer of discourse within the fictional world. In third-­person narrative, by contrast, there is supposedly only one source.2 Should this asymmetry be regarded as a strength, or as a weakness, for ON? While the point of the theory is to distinguish first-­person from third-­person narration, this distinction should not happen at the cost of losing sight of what is common to both kinds of fiction, as opposed to nonfictional narration. A closer look reveals, however, that there is no real duplication of sources in first-­person narration: the author is the material (and creative) source of the discourse, but the first-­person narrator is the logical, or assertive source, the one that holds epistemological responsibility, since the first-­person narrator presents the narrative information as facts, while the author does not. To put this differently, the author asks the reader to imagine, or make-believe, but the narrator presents the narrative as something to be believed. These illocutionary attitudes are distinct, but the text only bears the mark of the narrator’s testimony: fictional sentences typically say “p” rather than “imagine that p.” Even if fiction is conceived as a command to the imagination, fictional texts do not take the overt form of a command, because it would present the fictional world as nonexisting, and it would therefore prevent immersion.3 A MN theory restores the symmetry between the two kinds of fiction by attributing epistemological responsibility, that is, the telling as true fact to an abstract, or theoretical, narratorial function in the case of third-­person narration. My reasons for postulating a theoretical construct to fill the narratorial function in third-­person narration rest on the following arguments: (1) it allows a unified view of language-based fiction (which can be captured 2. An author who is not afraid of this asymmetry is Gregory Currie. He claims that in all narrative, there is an external author who is also the narrator: this is Conan Doyle for the Sherlock Holmes stories. In first-­person narrative, there is in addition an internal author-narrator, represented in Sherlock Holmes by Dr. Watson. Third-­person narratives, by contrast, have only an external author-narrator instance (2010, 65–67). Since this is also the case for nonfictional narratives, how then can they be distinguished from third-­person fiction? 3. I remember a fictional text that takes the form of a command to the imagination. It is a story by Dr. Seuss that goes like this (accuracy not guaranteed): “Think up a green sky; think of bloogs blowing by” (and so on for a page or two). The effect is clearly different from writing “The sky was green and bloogs were blowing by.” In one case, the reader imagines another world from the perspective of the real world; in the other, the reader imagines the fictional world from within this world.

64  •  C hapter 3

by Genette’s formula A  N)4 rather than splitting it into two kinds; (2) it relieves the author of the responsibility of fulfilling the felicity conditions of the textual speech acts; and (3) insofar as the theoretical construct can present variable degrees of individuation, from the zero degree of the camera-eye narrators of Hemingway to the fully individuated narrator of The Stranger, it does not presuppose a radical break between narrator-based and no-narrator fiction, and it is therefore better able to deal with weakly individuated narrators. Take, for example, the case of the narrator of Thomas Mann’s The Magic Mountain: while we learn nothing about [its] person, this narrator occasionally showcases [its] narrative activity through comments like these: “If we are to get on with our story, however, it will be impossible to fathom and analyze all of [Hans Castorp’s feelings]” ([1924] 1996, 207). Or: “The word ‘relationship’ is Hans Castorp’s, we refuse any responsibility for it” (1996, 138). What would ON theory do with such a narrator, who only intervenes to comment on [its] narrative performance? Is Magic Mountain a case of narrator-based or narratorless narration? In a MN theory, this narrator is weakly individuated, and there is no need for the reader to imagine [it] as a person engaged in a particular situation and telling the story for a particular reason; but there is nothing wrong with reflecting on the narrator’s storytelling decisions, which are also those of the author, because the narrator’s lack of individuation prevents [its] distancing from the author. When a third-­person narrator deprived of physical features peppers the story with comments, opinions, and general statements, [it] acquires some kind of individuality. This explains why some readers imagine third-­person narrators as gendered individuals, but in so doing they push their mental representation of the narrator beyond what is mandated by the text. An ON theory, by, contrast would be much less tolerant of such imaginings because it drives an ontological wedge between existing and nonexisting narrators. I also think that the fluid conception of narrator that comes with the postulation of variable degrees of individuation, from zero to full, is better able to deal with the phenomenon of morphing narrators than a theory based on a strict opposition between narrator-based and narratorless narration. By morphing narrators I mean such cases as the narrator of Flaubert’s Madame Bovary, who is first presented as a classmate of Charles Bovary but then turns into a classic third-­person narrator able to represent Emma’s private thoughts. An even more striking example of morphing is the narrator of Proust’s In Search of Lost 4. ON theories could satisfy this formula vacuously: if there is no N, then obviously A  N.

N arrator  •  65

Time, who alternates between first-­person homodiegetic narration and third-­ person heterodiegetic (for instance, in Swann in Love [Un amour de Swann], and in the episode narrating the death of Bergotte, who was not with the narrator Marcel when he collapsed in front of the Vermeer’s “View of Delft.”).5 In an ON theory, the text alternates between narrator-based and narratorless passages, and a sharp distinction must be made between them, whereas a MN theory admitting variable degrees of individuation allows the question “who is speaking?” to be backgrounded for large stretches of text, so that the reader will not notice the occasional disappearance of the first-­person narrator. In narratology as well as in science, theories must be judged by their power of prediction. In her book Le Narrateur: Un problème de théorie narrative, Sylvie Patron discusses my proposal to attribute third-­person narration to an impersonal, disembodied narrator (as presented in Ryan 1981), and she categorically declares it to be false (2020, 117–18). But after criticizing it for alleged inconsistencies and for misunderstanding speech act theory, she endorses the following conclusions of mine, which are all supported by my proposal of a “thin” narrator in third-­person narration: Possibility (though not necessity) of recognizing first-­person narration as the imitation of a genre of nonfictional narrative // impossibility of doing so in third-­person narration. Possibility of judging negatively the narrator’s storytelling performance, without extending this judgment to the author’s performance in first-­ 6 person narration // impossibility of doing so in third-­person narration. Possibility of interpreting the style or idiolect of the text as distinct from the author’s in first-­person narration // impossibility of doing so in third-­ person narration. Possibility of regarding the first-­person narrator as unreliable // impossibility 7 of doing so in third-­person narration. 5. The episode could have been told to Marcel, and he elaborates on it; but the passage gives far too strong a sense of direct presence to pass as a second-hand narration. 6. For instance, considering Tristram Shandy a lousy narrator but Laurence Sterne a brilliant novelist. 7. This argument presupposes that narratorial unreliability consists of the defense, by a narrator, of values that the author does not want the reader to endorse, or of a narrator’s blatantly mistaken assertion of facts. On the other hand, I do not regard as unreliable those narrators who trick the reader into false beliefs and who later reveal these beliefs to be false (a strategy found, for example, in “An Occurrence at Owl Creek Bridge,” by Ambrose Bierce, where the reader is made to regard subjectively hallucinated events as objective facts). These narrators act deliberately, whereas truly unreliable ones do not.

66  •  C hapter 3

All these distinctions derive from the fact that the first-­person narrator is individuated, and is therefore distinct from the author, whereas the third-­ person narrator’s lack of individuation prevents its distantiation from the author, except when it comes to presenting the narrative facts as true. If my conception of narrator leads to correct conclusions, can it really be regarded as false? In narratology as in science, different theories can be considered valid if they explain the same facts adequately. (Patron’s list includes three more items; see Ryan [1981] for a complete overview.) The proposal I have outlined is technically neither ON nor MN, if by MN one understands a naturalizing conception of narrator. ON shares with naturalizing theories of fiction the view that there is only one kind of narrator: an individuated, embodied, anthropomorphic member of the fictional world who possesses a distinct identity and engages in an act of communication that reproduces a genre of nonfiction. When no such figure can be identified, ON tells us that there is no narrator, while MN theories postulate one at any cost. My proposal opens an alternative way to be a narrator: to narrate can mean to be an agent of narration who presents a story as true of the fictional world while lacking the individuating features that define a person. Moreover, as the next section will show, there are many different functions that we associate with narratorhood. But I credit ON theory for liberating fictional narrative from the excesses of naturalization and narratorial personification, especially when this personification leads to the postulation within the most ordinary storyworlds of an individual gifted with the supernatural ability of omniscience.

Narratorial Functions What kind of involvement with language, the self, truth, and the storyworld does the narrative act involve? The rarity of the term narrator in everyday language suggests that in its nontechnical usage it occupies an uncertain territory overlapping with those of author, storyteller, actor, speaker, or writer. In narratology, by contrast, the role of the concept is well defined: narrator is used in opposition to author, and the disjunction author-narrator is, according to Genette (1997, 73), constitutive of fictionality. But the narrator should not be an ad hoc theoretical construct added on to the author parameter for the sole purpose of distinguishing fiction from nonfiction, for if this were the case we could greatly simplify matters by stating that nonfiction is narrated by an a and fiction by an n. By virtue of Occam’s razor, the postulation of both an author and a narrator in the case of fiction can only be justified if each of

N arrator  •  67

these parameters fulfills a distinct role that complements the function(s) of the other. Outside the territory of strictly literary fiction, the main usage of narrator is the designation of the oral performer of a narrative text in a multimedia environment, such as a story embedded in a musical composition (Prokofiev’s Peter and the Wolf), a film (voice-over narration), or a TV documentary (The Civil War, narrated by Ken Burns). On the other hand, when the story being orally performed is a joke, an improvised tall tale openly presented as a lie, or a traditional tale, we prefer the term storyteller. This suggests a discrepancy between oral and written brands of fiction: the narrator narrates the written fictional text; the storyteller performs the oral species. It seems somewhat artificial to speak of a contrast between author and narrator in oral storytelling, not only because authorship is a problematic concept when the performance involves traditional materials but also because the corporeal presence of the human source of language hinders the construction of an imaginary narratorial figure. If a bard, let’s call him Homer, engages in a public storytelling performance, the words obviously flow out of Homer’s mouth, and it appears counterintuitive to postulate an abstract narrator different from the performer. But in written communication, the creator of the tale is corporeally absent; this makes it easier to attribute the act of narration to an imaginary narrating voice whose existence is strictly a matter of make-believe. Interestingly enough, when the term storyteller is metaphorically used in discourse relating to literary fiction, it refers to the author (“a natural storyteller,” as many blurbs proclaim) and not to the narrator. The term storyteller connotes an invented tale, performed for the sake of entertainment and “displayed” to the audience (Pratt’s [1977] term), while narrator carries the expectation of a story told as truth, even when this is not literally the case, as in fiction. This telling, however, is not necessarily done in an artful nor a cooperative manner, as William Labov’s examples of reluctant narrators suggest. What, then, are the pragmatic functions, or responsibilities, involved in an act of narration? If we regard the conversational narrator of personal experience—that is, the instantiation of Labov’s (1972) “natural narrator” as the fullest form of narratorhood—these pragmatic functions are three: the creative, the transmissive (or performative), and the testimonial (or assertive). The creative function resides in the activity of shaping and encoding the story, and it presents therefore both a productive and a self-expressive dimension.8 While the creator produces the text (usually by writing it), it is not always the 8. It is within the creative function that I place the three narratorial functions distinguished by Phelan and Rabinowitz (Herman et al. 2012, 24): report, evaluate, explain.

68  •  C hapter 3

creator who communicates the story to an audience; hence the need to distinguish the creative from the transmissive function. The third function consists of presenting the story as true of its reference world, which means accepting responsibility for the truth of the propositions asserted by the narrative discourse. This function is admittedly more controversial than the other two: nobody will seriously deny that it takes the creation and transmission of discourse to tell a story, but as the case of the oral storyteller suggests, stories can arguably be told as invention. Yet there is always an interpretive level on which the audience of both a written narrative and an oral performance regards the text as a largely accurate statement of facts about a storyworld; otherwise, the storyworld would be in a state of such radical ontological indeterminacy that it would be impossible to construe a fabula, or story, made out of existents, states, and events. The creative, transmissive, and testimonial functions are thus the building blocks of the act of narration, but they can be performed by different agents. Table 1 shows the various possible modes of participation in a narrative transaction for a given individual in a real-life situation; when a slot is occupied by a minus sign, another narrating agent must step in and complete the script. A row with both plus and minus signs exists, therefore, in a complementary relation with another row. When the minus sign concerns the testimonial function, the complementary row appears in table 2. Row 1: Obviously, no narration take place if nobody assumes any of the three narratorial functions. Row 2: Here the person who fulfills the testimonial function is not a narrator but a witness. This row needs therefore to be complemented by another— typically row 8, in which another agent performs the creative and transmissive functions. Note that this other agent also performs the testimonial function: while each function must be fulfilled by somebody, the model does not prevent the duplication of functions. Row 3: The narrator transmits the text to an audience, but he or she is not the creator, and does not take responsibility for its truth. The creative function is fulfilled by the narrator figure of row 5. Row 4: The narrator both transmits the text and takes responsibility for its truth, but somebody else created it. I assume that TV broadcasters and narrators of documentary texts tacitly endorse the text as true, even though they are only performing it. This row is complemented by row 6 for the creative function. Row 5: This is the creative counterpart of row 3: the playwright needs actors to perform the text. This row also describes the situation of the author of a fic-

N arrator  •  69

TABLE 1. The distribution of narratorial functions in actual agents ROW CREATIVE

TRANSMISSIVE TESTIMONIAL

EXAMPLE

1

–

–

–

[No narration]

2

–

–

+

Witness vouching for the truth of somebody else’s testimony in a trial

3

–

+

–

Actor performing a narrative monologue or telling a story as part of a role; narrator of Peter and the Wolf

4

–

+

+

Well-known broadcaster narrating a TV documentary; religious figure taking dictation from God

5

+

–

–

Writing a play or a story to be performed orally by somebody else

6

+

–

+

Writer of a TV documentary

7

+

+

–

Oral storyteller (of tales or jokes); novelist

8

+

+

+

Historian, narrator of personal experience

tional story meant to be performed orally by somebody else, as in the case with the text of Peter and the Wolf. The minus sign in the testimonial column shows that the text is fictional. It will take a row of table 2 to fill out this function. Row 6: The writer of the text of a documentary film or of TV news also needs somebody else (row 4) to perform it. Row 7: The oral storyteller performs the text, and insofar as the text is improvised rather than read or memorized from a written source, he or she also functions as creator. The novelist writes the text and fulfills the transmissive function by publishing it. But insofar as both storyteller and novelist present their text as fiction, they do not vouch for its truth. This row must be complemented by a row of table 2. Row 8: Historians write the text and transmit it through publication. Narrators of personal experience improvise the text and fulfill the transmissive function through their public performance. Both present the story as true. Table 2 proposes a typology of fictional narrators based on the functions fulfilled by the narrating agent. In this table, the creative function is reinterpreted as self-expressive. A negative value for this function means that the narrative discourse is not felt by the reader as springing from the mind of an individuated, anthropomorphic being and does not express a narratorial personality. A negative value for the transmissive function means that the narrator is not imagined as “speaking,” “writing,” composing in a rec-

70  •  C hapter 3

ognizable genre, nor indeed as engaging in an act of communication that counts as an event in the history of the storyworld. And, finally, a negative value for the testimonial function means that the fictional enunciator is not really narrating anything but is simply mumbling or rambling, producing discourse without representational intent. In contrast to table 1, where values are reasonably binary, the plus and minus signs should be taken here not as absolutes but as coefficients susceptible of degree. Because the fictional narrator is a virtual entity, not a real-world individual like the agents of table 1, its properties are inferred by the reader, and they remain to some extent a matter of personal interpretation. Moreover, I conceive the values of narratorial functions as potentially fluid and renegotiable. A narrator who first appears to be a historian or memorialist may, for instance, fall out of the genre by resorting to mind-reading techniques; or, a narratorial personality may emerge and disappear in third-­person narration, destabilizing through this oscillation the value of the creative-expressive function. I illustrate the various rows in table 2. Row 1: In a narrator-optional model, this describes the situation of a narrative without a narrator: a third-­person narration that contains statements that could not be uttered in nonfictional communication because their felicity conditions cannot be satisfied in the real world. In a narrator-mandatory model, this combination is nonexistent. Row 2: In a narrator-mandatory model, this is how third-­person narratives without an individuated, embodied narrator are described. A good example is Hemingway’s “Cat in the Rain,” a text with camera-eye focalization that avoids any narratorial self-representation and therefore clearly deserves a minus in the creative category. More difficult to classify would be an intrusive third-­ person narration whose otherwise impersonal narrator manifests its presence through comments, opinions, judgments, generalizations, and addresses to the reader (such as Balzac’s narrators). To the extent that these interventions outline a personality, they swing the text toward row 6, but without achieving full narratorial individuation. Row 3: I do not know of any existing example of this combination. It would be fulfilled by a fictional editor character who presents to the reader a collection of fictional texts. This character does not present them as true (hence minus for testimonial) and is not their author (hence minus for creative/self-expressive). Row 4: This combination is represented by the editor figure of epistolary novels, who presents the letters as authentic. Like the hypothetical example of row 3, it presupposes an additional narrative layer supported by its own

N arrator  •  71

TABLE 2. The distribution of narratorial functions in fictional agents CREATIVE ROW (=SELF-EXPRESSIVE) TRANSMISSIVE TESTIMONIAL

1

−

–

EXAMPLE

–

Narratives without narrator according to the narrator optional theory: Hemingway’s “Cat in the Rain”

2

–

–

+

Third-person effaced narrator according to the narrator mandatory theory: Hemingway’s “Cat in the Rain”

3

–

+

–

?

4

–

+

+

The editor character of an epistolary novel

5

+

–

?

Molly Bloom’s monologue in Ulysses

6

+

–

+

J. M. Coetzee’s Waiting for the Barbarians

7

+

+

–

Scheherazade’s storytelling

8

+

+

+

Roquentin in Sartre’s La Nausée; Jane Eyre in Charlotte Brontë’s eponymous novel.

narrating agents, the letter-writers, who are typically described by the values of row 8. Row 5: Here we have narrators who create a discourse in their mind that expresses their personality (a plus for creative), but this discourse does not represent facts, and these narrators do not voice it or write it down. An example could be the monologue of Molly Bloom in James Joyce’s Ulysses. The text places the reader directly in the theater of Molly’s mind, without clearly distinguishing memories from day (or rather night) dreaming, and while we can infer that “Molly imagines / thinks of / dreams of p,” we have no basis for assuming that p holds in the storyworld. Row 6: This row describes the situation of a well-individuated first-­person narrator whose discourse may be taken both as a representation of facts and as the expression of a personality, but because of pragmatic impossibilities, this discourse cannot be naturalized as an act of communication. The text is therefore taken as the artificial transcription of a discourse that never leaves the mind. As Cohn has shown (1999, chapter 6), in J. M. Coetzee’s Waiting for the Barbarians, the narrator’s sustained use of the present implies living and

72  •  C hapter 3

telling at the same time—an illocutionary paradox showcased in the sentence “I doze and I wake” that prevents communication. This combination of values can be extended to any first-­person narrative that is not presented as being either written or orally told (The Stranger, by Camus). Row 7: This row presents the same combination of feature as row 7 of table 1, where this combination was described as constitutive of the author of a fiction. Therefore, it describes the situation of a fiction within a fiction, such as the discourse of Scheherazade in The Thousand and One Nights, and it introduces an additional level of fictionality. Row 8: This combination represents a conception of the narrator that became popular in the late 1970s, when literary critics discovered speech act theory and tried to use it as a basis for a characterization of fictionality. According to this characterization, whose main proponents were Mary Louise Pratt (1977) and Barbara Herrnstein Smith (1978), fiction is an “imitation speech act,” and narrators engage in an act of communication that reproduces a genre of real-world communication: writing a diary for Roquentin in Jean-Paul Sartre’s La Nausée, or an autobiography for Jane Eyre in Charlotte Brontë’s eponymous novel. Let’s conclude this section with a few remarks about the theoretical implications of my proposal. First, by linking all three functions to aspects of language pragmatics, it presents narratorhood as a matter of verbal deixis. This does not mean that I deny the possibility of narration in visual or dramatic media but rather that I resist the obligatory postulation of a narrator in film, drama, and painting. If narrative is an interpretive model for verbal, visual, and perhaps even musical information, this model can be suggested to the experiencer by a variety of clues. The verbal presentation of a story by a narrator is the most efficient way to activate a narrative interpretation but by no means the only one. Second, the so-called narrators of the examples are not all narrators in a literal sense, that is, agents in the production of narrative meaning. Some of the examples, such as row 5 of table 2, involve texts of very marginal narrativity. It is by convention that we attribute these texts to narrators, even though the texts they present do not fulfill the basic conditions of narrativity. Third, by dividing narratorhood into three distinct functions, the model questions the universality of the theoretical correlate of the narrator, namely the narratee. The concept of narratee is strongly implied by function 2, weakly implied by function 3, and not implied at all by function 1. The degree of presence of the narratee is therefore as variable as the visibility and mode of existence of the narrator.

N arrator  •  73

And, finally, by permitting a whole range of ontological realizations for the narrator, my proposal smooths out the uneasy relationship between classical narratology and its “natural” cousin, an approach inspired by discourse analysis which provides fresh insights into literary narrative but tends to reduce fiction to an imitation of conversational and historical genres. The present model secures a room in the house of literary narratology for embodied fictional narrators who communicate in a simulacrum of reality-based discourse, but it reminds us that the narrator is a theoretical fiction, and that the humanlike, pseudonatural narrator is only one of its many possible avatars.

CHAPTER 4

Characters Textual, Philosophical, and “World” Approaches to Character Ontology

During the years when the study of literature was dominated by a global movement that I call textualism, whose individual manifestations include New Criticism, deconstruction, and postmodern theory, it was common when speaking of literary characters to oppose a “naïve reader” who regarded characters as persons, or as model human beings (Herman in Herman et al. [henceforth NT] 2012, 125), with an enlightened reader who knew that characters are constructs made of language. The following declarations are meant by their authors to uncover the “true” essence of characters: “Characters in fiction are, after all, words on a page” (Richardson in NT 133). “Characters are marks on a page, made of the alphabetical characters that spell out ‘who’ they are. They have no psychology, no interiority, no subjectivity” (Warhol in NT 119). The epitome of the naïve reader is Don Quixote (part 2, chapter 26), who jumps on stage to rescue the heroine of a puppet show, causes a huge brawl during which all the puppets are broken, and ends up having to pay for the damage (and recognizing his error). We all know that Don Quixote was crazy, his brain having dried up from reading too many chivalry novels, and that he was unable to distinguish fiction from reality. But there are many people who engage in behaviors that orthodox literary theory would consider naïve: for instance, kids watching puppet shows who scream to warn the hero that the bad guy is approaching; people who travel to Reichenbach Falls in Switzerland to see the spot where Sherlock Holmes fell to his death (only to be resurrected 74

C haracters  •  75

later); and especially people who cry when reading fiction: Dickens’s Little Nell is reported to have let loose torrents of tears. Are these people as crazy as Don Quixote, or do they know something about the nature of fiction and about characters that the puritans of textualism refuse to acknowledge? In this chapter I try to answer this question by exploring and contrasting three approaches to character: textualist, philosophical, and the approach that I personally endorse, which I call the world approach.

Textualism Textualism bears prime responsibility for claims that characters are not persons but objects made of language. The movement grew out of structuralism, and it is a matter of common knowledge that structuralism drew inspiration from linguistics, especially from the linguistic theory of Ferdinand de Saussure, who regarded language as a self-enclosed system rather than as a means to represent an external reality. Under the influence of Claude Lévi-Strauss, linguistics was revered by structuralists as a pilot discipline, and the adoption of its methods in other fields such as anthropology, sociology, and, last but not least, literary theory was considered a way for these disciplines to reach scientific status. The big question, however, was how to apply linguistic methods to other domains of signification, for beyond the concept of sign and its division into a signifier and signified, and beyond the claim that the value of signs depends on their systemic relations with each other rather than on their relation to the world, Saussurian linguistics did not provide particularly concrete directions. The adoption of a linguistic model in the humanities and social sciences was mainly metaphorical and ideological: metaphorical because it consisted of regarding every phenomenon under study as a “language” based on a “code” made of discrete signifying units, and ideological because it viewed human thought as profoundly shaped by these multiple codes. In its most radical conception, the linguistic influence means that we are spoken by language rather than speaking it. The same could be said about the various codes of culture: even nowadays, cultural studies have a strongly deterministic bent. It is against this background that I would like to discuss the treatment of literary characters by the most famous of structuralist critics, Roland Barthes, in S/Z (1970; English 1974), his well-known study of a short story by Balzac titled “Sarrasine.” The ambition of Barthes in S/Z was to promote a view of literary texts as systems constituted by multiple codes whose interaction determines meaning in an endless play of signification. But this play can be more or less extensive; in the type of text he calls the writerly [scriptible], signs activate each other

76  •  C hapter 4

in all directions, there is no beginning nor end, and the reader becomes a producer of meaning; in the type of text he calls the readerly [lisible], the play of signification is limited by the linearity or sequentiality of narrative structure and by the demands of mimeticism. But even for readerly texts, Barthes opposes the classical literary-critical position that regards meaning as imposed top-down by the author. While he conceives the writerly text as fully created bottom-up by the reader, the readerly text represents a limited plurality where bottom-up interpretive activity is at least partly controlled by top-down structures. Yet Barthes does not go as far as saying that these structures are imposed by an author: instead, they are the product of cultural codes and of language itself. It is interesting to note that despite praising the writerly as the future of literature and as the liberation of the reader, Barthes did not devote nearly as much attention to such texts as he did to the readerly kind. It could be because the writerly does not really exist; but I rather believe that, deep down, Barthes preferred the readerly because the readerly preserves narrative interest, which lies at least in part in temporal effects such as suspense, curiosity, and surprise (Sternberg 1992). Barthes regarded narrative texts as being regulated by five codes: the proairetic, which organizes the actions of characters into meaningful sequences, such as “taking a walk” or “rendez-vous” or “assassination”; the hermeneutic, which organizes narrative information into the presentation and then solving of an enigma; the semic, which consists of gathering the connotations of textual units, for instance extracting “wealth” from the description of a fancy house; the symbolic, which links particular existents to universal concepts; and the cultural, which links textual units to established forms of knowledge, especially popular wisdom and stereotypes. The designation of these interpretive moves as “codes” betrays the dominance of the linguistic paradigm; nowadays we would be more inclined to regard them as mental operations, and instead of ascribing the functioning of the semic code to codified relations between signifiers and signifieds, we would recognize the importance of the reader’s life experience for extracting connotations and making inferences. But whether or not Barthes’s codes are really codes in the strict sense of the term, each textual unit is justified by its participation in one or more of them; the more numerous the codes, the more determined and meaningful the textual unit. These units can be segments of variable length, and their delimitation is determined not by the kind of systematic discovery procedure that structuralist linguistics made into their number one priority1 but by whether Barthes had something to say about them or not. 1. An example of such a procedure is the replacement of one sound by another in a word to show that they contribute to meaning distinction and are therefore distinct phonemes rather

C haracters  •  77

Like any other textual unit, characters lie at the intersection of several codes. Take Zambinella, the castrated male singer with whom the artist Sarrasine falls in love, believing that he is a woman. According to the symbolic code, Zambinella represents superfemininity and submasculinity. The hermeneutic code presents her sexual identity as an enigma to be solved. According to the cultural code, she stands for ideal beauty. The proairetic code enables the reader to gather her actions into meaningful sequences, such as a sequence of “playing a trick on Sarrasine by pretending to be a woman,” and the semic code enables the reader to extract the connotations of the words that describe her, for instance, linking her mouth to sensuality. More generally, for Barthes a character is a collection of semes (that is, meanings) subsumed under the heading of a proper name. Just as the proairetic code instructs the reader to gather information under a general type of action such as “murder,” “walk” or “rendez-vous,” the proper name unifies a dispersed collection of semes and turns them into a character: Sarrasine is the sum, the point of convergence, of: turbulence, artistic gift, independence, excess, femininity, ugliness, composite nature, impiety, love of whittling, will, etc. What gives the illusion that the sum is supplemented by a precious remainder (something like individuality, in that, qualitative and ineffable, it may escape the vulgar bookkeeping of compositional characters) is the Proper Name, the difference completed by what is proper to it. The proper name enables a person to exist outside the semes, whose sum nonetheless constitutes it entirely. (1974, 191; italics original)

The last sentence in this quote shows that Barthes’s conception of characters is more complex than reducing them to a “collection of semes.” The proper name does indeed turn characters into persons. But where do these persons exist, if the text is entirely constituted of semes? The only answer can be this: they exist as persons in the storyworld as well as in the imagination of the reader. But this existence is an illusion because, as Barthes notes, the semes “constitute them entirely.” The instrument of this illusion is the proper name, which suggests the existence of a referent external to language. Barthes admits that his own critical discourse is not always immune to the illusion of existence created by the proper name: We occasionally speak of Sarrasine as though he existed, as though he had a future, an unconscious, a soul; however, what we are talking about is his than mere variations in pronunciation. Thus, [b] and [p] are considered distinct phonemes in English because “bit” and “pit” form a minimal pair with different meanings.

78  •  C hapter 4

figure (a impersonal network of symbols combined under the proper name “Sarrasine”), not his person (a moral freedom endowed with motives and an overdetermination of meaning): we are developing connotations, not pursuing investigations; we are not searching for the truth of Sarrasine, but for the systematics of a (transitory) site of the text. (94)

In other words, the proper name is deceptive because it presents characters as persons and suggests that they exist independently of the text. The task of the critic is to demystify the text, to denounce the illusion, and to prepare the advent of the “writerly text,” which will do away with illusion: “What is obsolete in today’s novel is not the novelistic, it is the character; what can no longer be written is the Proper Name” (95). Barthes wrote in an age when the advocates of the New Novel claimed that the novel had to do away with such fundamental narrative elements as plot and character (Robbe-Grillet 1965a, 25–47). To empty characters of any human substance, it often replaced proper names with bare initials: for instance, the wife in Robbe-Grillet’s La Jalousie is referred to as A. But the New Novel was a short-lived literary fashion: today’s novels are still full of characters, and the proper name remains the main way of referring to them. Barthes’s prophecy turned out to be dead wrong. To demonstrate that the behavior of characters is not entirely dependent on the realism of psychological motivation, Barthes discusses a passage where Zambinella tries to confess to Sarrasine that she is not a woman but a castrated male. The text goes like this: “Listen, monsieur,” she said in a low voice. “Oh, be still,” the impassionate artist said. “Obstacles make my love more ardent.” (177)

As a result of being told to shut up, Zambinella does not make the confession. Here is Barthes’s comment: If we have a realistic view of character, if we believe that Sarrasine has a life off the page, we will look for motives for this interruption (enthusiasm, unconscious denial of the truth, etc.). If we have a realistic view of discourse, if we consider the story being told as a mechanism which must function until the end, we will say that since the law of narrative decrees that it continue, it was necessary that the word castrato not be spoken. (178)

If Zambinella had been able to make the confession, the story would not have reached its dramatic climax, the murder of Sarrasine by Zambinella’s protector

C haracters  •  79

during a later meeting. What Barthes calls the demands of discourse are the demands of plot, and they override the demands of verisimilitude: Sarrasine’s refusal to hear Zambinella’s confession is an ad hoc response that cannot be justified on psychological grounds.2 Here Barthes confronts a dilemma that is often invoked by theoreticians of character: should characters be subordinated to plot—in which case they are mostly cogs in a system—or should plot be subordinated to character? The answer depends on genre: according to Aristotle, in tragedy character is subordinated to plot (Poetics 4.3; 1996, 11);3 in contrast to tragedy, one can assume that in epics, plot is subordinated to character. The purpose of the multiple episodes of the Odyssey can, for instance, be said to be the demonstration of the personality of Odysseus. But in some cases, psychologically or pragmatically motivated actions by characters would prevent an interesting development. To avoid this pitfall, authors often sacrifice credibility to the demands of plot. A common example is the convention of the calumniator credited (Steinmann 1981, 258), where an intelligent character believes the lies of a character of low reliability, such as Iago in Shakespeare’s Othello, because this leads to a crisis which is central to the plot. If characters are nothing more than collections of semes, there is nothing to lose by accepting their subordination to plot; but Barthes’s stance is more complex, as we see from this passage: “From a critical point of view, therefore, it is as wrong to suppress the character as it is to take him off the page in order to turn him into a psychological character (endowed with possible motives): the discourse and the character are each other’s accomplices” (178). Here Barthes is telling us that it is wrong to assume that characters are human beings and that they act out of psychological motivations but that it is also wrong to suppress characters in the name of discourse. In other words, Barthes wants to have his cake and eat it too. He escapes this impasse through a magic trick, typical of textualism, by refusing to make a distinction between characters and discourse: “The characters are types of discourse, and, conversely, the discourse is a character like the others” (178). Like the others—really? When Barthes writes that “the discourse is a character like the others,” he wants to redirect the reader’s attention from the actions of characters to the action of the narrative discourse itself, to the twists and turns this discourse takes, and to the strategies it deploys to achieve its goal of proper narrative form. Naïve readers read texts as being about the lives, the adventures, and the schemes of characters; sophisticated readers read texts as being about the lives, the adventures, and the schemes of discourse. 2. It is therefore what I call a plot hole in chapter 5. 3. Throughout this book, I refer to Aristotle’s Poetics through both the section number (which is independent of particular editions) and through the page number of the edition I use.

80  •  C hapter 4

Barthes’s more or less implicit goals, in writing S/Z, were aesthetic and pedagogical. He wanted to promote a new mode of reading that broke with the biographical and positivist tradition that dominated French academia early in his career, a mode that liberated textual energies and that allowed readers to find pleasure in playing with language. He also wanted to promote an alternative to mimeticism and realism, an alternative embodied in the writerly, though deep down his loyalties remained with the limited pluralism of classical narrative4 rather than with the chaotic multiplicity of the writerly. With regard to characters, he seems to have realized the limitations of a purely semiotic approach, and he opted for ambiguity rather than openly admitting that there is more to characters than collections of semes.

The Philosophical Approach5 Philosophers, especially those of the analytic school, have been interested in fictionality since long before narratologists paid attention to it—which is a fairly recent development. For analytic philosophy, the questions that need answering are very different from Barthes’s concern with the functioning of literary language. These questions are the following: • How can we refer to fictional characters and make statements about them, for instance by saying “Anna Karenina is a more passionate lover than Emma Bovary”? • How does one assess the truth of statements made by critics about fictional characters? • Are there relations of identity between, say, Marlowe’s Faust and Goethe’s Faust, or are they mere homonyms? • And, last but not least, what is the mode of existence, or ontological status, of fictional characters? Do they represent a special mode of being, a position that transgresses a principle dear to philosophers, namely Occam’s razor, or should one say that they do not exist at all?

4. Barthes regarded narrativity as an obstacle to pluralism: “All of which comes down to saying that for the plural text, there cannot be a narrative structure, a grammar or a logic; thus, if one or another of these are sometimes permitted to come forward, it is in proportion (giving this expression its full quantitative value) as we are dealing with incompletely plural texts, texts whose plural is more or less parsimonious” (1974, 6; italics original). 5. Since the world approach discussed below is inspired by possible worlds philosophy, the label philosophical approach should be understood as relying on a one-world philosophy.

C haracters  •  81

The question of the ontological status of fictional characters is deeply entwined with the question of reference because reference is widely believed by analytic philosophers to imply some kind of existence. According to Bertrand Russell, a sentence about a nonexisting entity such as Emma Bovary is necessarily false because of referential failure; according to Gottlob Frege’s more flexible account, it is not false but indeterminate. But in everyday life we frequently refer to fictional characters, and we intend our statements to say something true: for instance “John is a real Scrooge”; or “Bill has seduced more women than Don Juan.” Moreover, if it were impossible to make true statements about fictional characters, the only criterion of validity for the claims of literary critics would be whether they are provocative enough to excite passions. Here I focus my discussion of philosophical approaches on the theory of Amie Thomasson, as developed in her 1999 book Fiction and Metaphysics. Thomasson believes that fictional characters can be the target of reference. They must therefore have some mode of existence. But this mode of existence is not the one of concrete, material objects such as people and apples, which are located in space and time, nor the one of purely abstract entities such as numbers or beauty, which exist eternally and cannot be traced back to any specific human creative act. For Thomasson fictional characters are what she calls “abstract artifacts,” and as such they occupy a middle ground between the material and the mental, the concrete and the abstract. Characters are abstract because they do not exist in space and time, and they are artifacts because they depend on literary works as well as on the concrete existence of authors and books (or other media). “In short, on this view fictional characters are a particular kind of cultural artifact. Like other cultural objects, fictional characters depend on human intentionality for their existence” (1999, 14). Worth noting in this definition is the claim that characters depend on the intentional act of the author, not on the text itself, as textualists like Barthes would probably claim. If two authors happened to write the same words, unbeknownst to each other, they would produce distinct literary works and different characters.6 In Thomasson’s view, fictional characters are born through the creative act of the author, they are maintained in existence through the media that make them accessible to readers, and they die when the last copy of the work, or its memory in a reader’s mind, is destroyed. This idea that fictional characters can “die” may seem counterintuitive; isn’t “death” a phenomenon specific to concrete, material objects? And yet we say that languages die together with 6. Jorge Luis Borges played on this idea when he claimed, in his short story “Pierre Mesnard, Author of the Quixote,” that Mesnard’s Quixote is an entirely different work from the one by Cervantes, even though they match word per word.

82  •  C hapter 4

their last speaker; why should it be different with characters? If something can die, it means that it once existed.7 Through her concept of abstract artifact, Thomasson opposes theories that regard characters as unrealized possibilities. According to these theories, Hamlet could have existed, and the world could contain elves, trolls, and hobbits, as it does in Lord of the Rings. This conception of fictional characters as possible entities is supported by Aristotle’s claim that the task of the poet is to show “not what is but what could be in accordance with probability and possibility” (Poetics 5.5). But there are countless possibilities that have not been textualized. Possibilities exist independently of whether somebody thinks of them or not; therefore, if one takes the view that fictional characters are unactualized possibilities, then authors do not “create” or “invent” characters but instead “discover” them. Thomasson rejects this idea on the grounds that it makes fictional characters independent from the creative acts of authors. Yet if characters do not pre-exist the text as unactualized possibilities, if they are not “discovered,” this raises the question of how they arise in the author’s mind. Authors do not create characters ex nihilo; they make them up by mentally exploring a field of possibilities, and by selecting some of these possibilities to be realized textually. An important feature of Thomasson’s proposal is that it defines the ontological status of characters from an external point of view—that of the author who inhabits the real world rather than that of the narrator and characters who are internal to the story. If fictional characters are the product of an act of creation that takes place in space and time, then their properties are entirely determined by this act of creation. And since the author can only imagine and encode a limited number of properties, this means that fictional characters are ontologically incomplete. According to this view, the number of Lady Macbeth’s children, which remains unspecified in Shakespeare’s tragedy, represents an ontological gap, a hole that cannot be filled in Lady Macbeth’s character, and as the critic Lionel Charles Knights argued as early as 1933, it would be pointless to try to answer the question “how many children had Lady Macbeth?” The vast majority of the philosophers and literary critics who 7. Françoise Lavocat pursues this idea in a charming “theoretical fiction” titled Les Personnages rêvent aussi (Characters Also Dream, 2020). She imagines that fictional characters, who can come from many media, inhabit a planet called Fiction, but they live only as long as people on planet Earth are aware of their existence. Many literary characters are therefore threatened with extinction, especially in an age where reading is losing ground to film, TV, and video games. The endangered individuals include many of the heroes of Balzac’s Comédie Humaine, since only a few of the ninety-one novels and short stories that make up the work are still widely read today. Vautrin, the main villain of Balzac, has a brilliant idea to keep them alive: turn them into the heroes of video games.

C haracters  •  83

have addressed the problem of the nature of fictional characters endorse the idea of their radical incompleteness, for instance Uri Margolin, Thomas Pavel, Lubomír Doležel, Ruth Ronen, and Nicholas Wolterstorff. This incompleteness could be seen as what distinguishes abstract artifacts from real persons. But the thesis of incompleteness runs into problems when a member of the real world appears in a novel, for instance Napoleon in War and Peace. Should one postulate the existence of a real, ontologically complete Napoleon, as well as of a fictional, incomplete version of the emperor? If these two Napoleons are ontologically so different from each other, how can one explain that, for the reader, they are versions of the same person? Readers of War and Peace will imagine the fictional Napoleon as sharing many properties with the real one, even when these properties are not specifically mentioned in the text. For instance, if readers try to mentally visualize a battle scene of the novel where Napoleon appears, they are entitled to imagine Napoleon as short, putting his hand in his coat, riding a white horse, and wearing his hat sideways, as shown in the painting by Meissonier. I am not saying that they have to imagine him in such detail, but rather that there is nothing wrong in doing so. On the other hand, it would be wrong to imagine Napoleon as looking like Don Quixote. Thomasson does recognize the limitations of a theory that regards incompleteness as a defining ontological feature of fictional characters. She considers the two statements “According to the story Hamlet has blood type A” and “According to the story Hamlet has blood type B” as both false or indeterminate because Shakespeare’s text says nothing about Hamlet’s blood type (1999, 108). But it does not follow from the negative or indeterminate truth-value of these statements that “According to the story Hamlet is incomplete as to blood type.” The reason for the failure of the entailment is that, according to the story, Hamlet is not a fictional character but a regular human being. He is not created by an author but born of a father and mother. And since regular human beings are ontologically complete, one must assume that Hamlet is 8 also complete, at least within the story. The difference between an internal and an external point of view can be illustrated by these two statement types: “Hamlet is a fictional character who was created by Shakespeare” and “Hamlet is the Prince of Denmark.” The second statement can be paraphrased as “according to the story, Hamlet is the Prince of Denmark.” By adding the prefix “according to the story,” an internal statement can be turned into an external one and receive a positive truth 8. This is an application of what I called the “principle of minimal departure” in Ryan (1991, chapter 3). The principle (inspired by Lewis 1978) states that when we imagine fictional worlds and entities, we are entitled to project upon them everything we know about the corresponding entities in the actual world, unless contradicted by the text.

84  •  C hapter 4

value for the real world. But this operation fails with external statements: we cannot say that “according to the story, Hamlet is a fictional character who was created by Shakespeare.” Thomasson retains an external perspective when she characterizes fictional characters as abstract artifacts, for it is certainly not true that, according to Shakespeare’s play, “Hamlet is an abstract artifact.” In her model, some statements about fictional characters must be prefixed by “in the story” in order to be evaluated as true or false, and others should not be. But how do we know which ones should be prefixed? An alternative to “externalizing” internal statements with a prefix is to make the contrast between an internal and an external point of view the cornerstone of a theory of the ontological status of fictional characters. If we replace the prefix “in the story of Hamlet” with “in the world of Hamlet,” we can account for the duality of perspective through a theory inspired by the concept of possible worlds. It is to this theory that I turn next.

The World Approach The world approach to character is based on possible worlds (PW) theory, whose conception of fictionality I presented in chapter 2. Rather than repeating here the entire discussion, I limit my presentation to the ideas that make the theory particularly relevant to the question of characters. The first of these ideas, which is embraced by all PW philosophers (see introduction), is the contrast between an actual world, where “things are what they are,” and nonactual possible worlds, where things are different. Since things are what they are in only one way, whereas they could differ from the actual state in many different ways, it follows that there is only one actual world and many nonactual possible worlds. The other important idea is the indexical conception of actuality, proposed by David Lewis. As he writes: Our actual world is only one world among others. We call it actual not because it differs in kind from all the rest but because it is the world we inhabit. The inhabitants of other worlds may truly call their own world actual, if they mean by “actual” what we do; for the meaning we give to “actual” is such that it refers at any world i to the world i itself. “Actual” is indexical, like “I” or “here,” or “now”: it depends for its reference on the circumstances of utterance, to wit the world where the utterance is located. (1979, 184)

C haracters  •  85

Lewis’s indexical conception of actuality underlies a position known as modal realism, according to which all possible worlds exist objectively. This idea may be difficult to accept, but it works very well for fictional worlds and their inhabitants. If actual is indexical, fictional characters are actual, embodied, and ontologically complete individuals from the point of view of the fictional world, but from the point of view of our actual world, they are the abstract artifacts Thomasson describes. By referring to the characters as if they were real persons, and by describing their world as real, the fictional text invites readers to transport themselves in imagination into the fictional world, and to adopt the point of view of one of its members. I have called this mental transportation recentering (Ryan 1991, 18ff.), and I regard it as essential to the experience of immersion in a storyworld. The world account has multiple advantages, among them these: • The dual perspective makes it possible to take both the author and the narrator into consideration. The textualist perspective of Barthes eliminates both author and narrator; the philosophical perspective of Thomasson takes the point of view of the author. Here, characters are both made-up creatures from the point of view of the author and individuals who exist independently of the text from the point of view of the narrator. Their story is invented from the point of view of the author, but it is told as true fact from the point of view of the narrator. • Because storyworlds can be contemplated from both an internal and an external perspective—the perspective of the real world—the world model allows users to move back and forth between these two perspectives, and it explains how characters can be regarded as both human beings and as textual constructs. James Phelan (1989) has identified three so-called functions that characters can fulfill: the mimetic (passing as a person), the synthetic (reminding the reader that the character is fabricated), and the thematic (standing for an idea). The mimetic function represents the internal perspective; the synthetic and the thematic functions represent the external perspective. • The world model solves the problem of the presence of actual individuals such as Napoleon in fictional texts. If things could be different from what they are, as PW theory tells us, there are worlds where Napoleon has properties different from the Napoleon of the actual world, for instance worlds where he wins the battle of Waterloo, and worlds where he interacts with the heroes of Tolstoy’s War and Peace. The difference between Napoleon and Natasha is that Napoleon exists in both our actual world

86  •  C hapter 4

and in the world of War and Peace, whereas Natasha does not exist in our actual world. But within the world of War and Peace they share the same ontological status. • If real individuals can have counterparts in different worlds, so can purely fictional individuals. This explains the practice of transfictionality (SaintGelais 2011), or the phenomenon of fan fiction, which consists of writing stories that change some of the features of fictional characters and place them in different worlds but maintain a connection with the original manifestation of the character. Authors can also expand pre-existing fictional worlds, for instance by writing new stories about a character that fully respect the character’s original properties. • The world approach does not limit characterhood to human beings. The field of the possible is vast, and we can imagine worlds with species entirely different from those of our world: dragons, elves, fairies, witches, talking animals, Martians, and robots smarter than humans. All it takes to turn members of these species into characters is to give them distinctive mental attributes and cognitive abilities that turn them into agents: abilities such as free will, desires, and self-consciousness. • By ascribing to characters the same ontological status within their world as to human beings within our world, the world model easily explains emotional reactions to characters, such as empathy or intense dislike. We cry for Little Nell not because she represents an abstraction, as Richard Walsh has suggested (2007, chapter 8), but because we transport ourselves in imagination into a world where she exists as an innocent little girl who suffers undeserved hardship and dies an early death.

As we can see from this list of advantages, the world approach explains a phenomenon that neither the textual nor the philosophical approach is trying to address. This phenomenon is the reader’s experience of characters, or, to put it differently, the behavior of readers with respect to characters. The textual approach took an external perspective; it asked how readers assemble characters from scattered information, and it warned them against mistaking these textual constructs for persons. The philosophical approach was concerned with the possibility of making true or false statements about entities that do not exist in space and time, and it asked about the mode of being of these entities. This is not the kind of question that readers normally ask. The strength of the world approach lies in its ability to deal with behaviors that the textual approach regards as unworthy of an aesthetic appreciation of literature and that the philosophical approach regards as outside its field of expertise. By allowing a dual perspective, both internal and external, the world approach

C haracters  •  87

explains how characters can be experienced as both persons existing autonomously and as textual creations; how characters can both appear to act out of their own free will and be used to represent certain themes and ideas; how characters are both tied to their world and able to migrate to other texts and other worlds; and, last but not least, how readers can cry for characters while fully enjoying their crying, for unlike Don Quixote, readers are aware that characters are not real people. But what about characters who are not possible persons, because their properties are self-contradictory, because they lack the cognitive abilities that make it possible to interpret their gestures as meaningful actions, because they are flat allegories lacking any human substance, because they regard themselves as fictional and not real, or because they present ontological gaps that cannot be regarded as missing information? (I regard the characters of Beckett’s Waiting for Godot as embodying this kind of radical incompleteness.) In his work on unnatural narrative, Brian Richardson (2006, 2016) has presented many examples of characters who lack the dimension of personhood. The textual approach has no problem with them; they are just collections of semes gathered around a name, a pronoun, or an actor’s body in theater and film. The philosophical approach makes no categorical distinction between characters: they are all abstract artifacts created by authors. Can the world approach deal with such creations and distinguish them from fully developed characters? I see two ways of approaching this problem. First, we could say that characters lacking the status of possible persons are entities that block the world-internal perspective and limit the reader to the external stance. Blocking the world-internal perspective means preventing the imagination from filling these characters with anthropomorphic substance and forcing awareness of their synthetic or made-up nature. The only operations left to the reader are asking what abstract idea the characters are supposed to represent and how they contribute to the global organization of the text. The more strongly characters represent a theme or idea, the weaker their status as persons and their perceived autonomy: allegorical figures are the puppets of the author, not creatures who seem to act out of free will. Can one still call these creations characters? Only if one defines characters as the referents of proper names in fictional texts. A second approach would consist of retaining a conception of characters as possible persons and of denying some of the referents of proper names the full status of characters. There is no reason why every fictional text should create something worth calling a world; similarly, there is no reason why every fictional text should tell a fully developed story that involves individuated and ontologically complete persons. Characterhood, worldness, and narrativ-

88  •  C hapter 4

FIGURE 3. Degrees of characterhood

ity are not binary features but scalar properties of the mental representations elicited by texts: there are texts of low narrativity (Ryan 2007b), there are texts whose semantic content does not really congeal into a world because it lacks logical coherence (see chapter 8), and there are referents of proper names that lack individuating and mental human substance. If there is such a thing as an “unnatural character,” it is not a fantastic creature representing a species that cannot be found in the real world, it is an entity that is not fully realized as a person, that appears in a text of low worldness and narrativity, and that belongs to the margins of the fuzzy set of characters. Figure 3 represents the various degrees of characterhood. I place on the outside characters who have no proper name, who are ontologically incomplete, who embody contradictory properties, who have no stable identity, who are mentioned but do not appear on the narrative scene (cf. Godot), or who exist only as unrealized possibilities, such as the wife Frankenstein did not make for his monster. At the center of the fuzzy set I have put Don Quixote, Elizabeth Bennet, Mr Darcy, and the Star Wars crew. These are characters whose behavior we

C haracters  •  89

can understand, even when we do not approve of their choices, because they have basically the same reasons for acting that we do. They happen to have inspired intense transfictional activity, such as transmedial adaptations, fan fiction, prequels, sequels, and transpositions. (As Darth Vader shows, they do not have to be possible members of the real world.) But Hamlet and Emma Bovary, Sherlock Holmes and Little Nell, Donald Duck and Tintin also belong in this inner circle because the fullest of characters are those that speak so strongly to the imagination that they live beyond their text, whether this life inspires transfictional developments or remains within the readers’ minds.

CHAPTER 5

Plot Cheap Plot Tricks, Plot Holes, and Narrative Design

In narrative, plot exists on two levels: the plotting of the author, who creates the storyline; and the plotting of the characters, who set goals, devise plans, schemes, and conspiracies, and try to arrange events to their advantage. The plotting of both author and characters is meant to exercise control: for the author, control over the reader, who must undergo a certain experience; for the characters, control over other characters and over the randomness of life. But sometimes the goals of the author are at odds with the goals of characters. The author needs to make the characters take particular actions to produce a certain effect on the reader, such as intense suspense, curiosity, or emotional involvement; but acting toward this situation defies narrative logic because it is not in the best interest of the characters, or not in line with their personalities. In this chapter I propose to investigate two types of contrived plot twists that arise from this conflict between author and character goals. One involves a positive intervention by the author, an attempt to fix the problem through hackneyed devices; I call these cheap plot tricks (henceforth CPT). The other results from ignoring the problem, or covering it up, a strategy (or omission) that leads to what is known among film writers as plot holes (henceforth PH). Through this emphasis on the kinds of event that make the sophisticated reader groan, I break away from the almost exclusively descriptive tradition of both classical and postclassical narratology, and I adopt an evaluative stance closer to the prescriptive spirit of Aristotle’s Poetics. 90

P lot  •  91

In his treatment of tragedy, Aristotle sketches a catalog of good and bad ways to construct plot. He posits as pivotal to the tragic plot two event types: reversal of fortune and scenes of recognition (anagnorisis), through which characters pass from ignorance to knowledge. (Plots are even better when both events occur at the same time, as in Oedipus Rex.) In Aristotle’s examples of inferior plotting, recognition is achieved by means of telltale objects and external tokens, such as Ulysses’s scar that reveals his identity, or is “contrived by the poet” (8.2; 1996, 26); in truly artistic plotting, recognition is brought by inferences drawn by the characters, by memory or, in the best case, arises “from the events themselves”: from actions that are probable within the circumstances set up by the plot. A prime example of such motivation is Oedipus’s decision to launch an investigation of the murder of his father and discovering, as a result, that he is the culprit. One can generalize from these examples that good plots are propelled by the inner disposition of characters and by their logical reasoning whereas bad plots are steered by ad hoc external circumstances which bear the stamp of the author’s fabrication. A CPT is an event that is poorly prepared, that looks forced, that seems to be borrowed ready-made from a bag of tricks, and whose function for the plot as a whole is too obvious; in short, it is a narrative cliché. This is why I call it a plot trick rather than a plot twist. The vulnerability of an evaluative stance lies in the subjectivity of the reader’s judgment. What I label a CPT you may find very acceptable. If what passes as cheap was entirely a matter of personal judgment, it would make no sense to attempt a taxonomy of CPTs: each type of event could be used for good or bad plotting depending on the reader’s opinion of the skills of the author. The opposite stance consists of saying that there are some kinds of events that represent bad plotting, no matter the context. A compromise can be achieved between the relativism of the first position and the essentialism of the second by regarding some plot twists as inherently cheap while recognizing that they can be redeemed by being put in the service of a good story. But the most favorable reaction these plot twists will elicit is to be judged “excusable”: to be assigned to the valleys and not the peaks in the contour of the plot. In the reader’s aesthetic evaluation, plotting devices range along a continuum from cheap to brilliant, with a middle occupied by events that do not provoke strong reactions. In this chapter I focus on CPTs and PHs rather than on brilliant plot twists (BPTs) because their identification, having to do with faulty logic, implausibility, or a sense of déjà vu is much less dependent on the reader’s personal taste, and they are therefore much easier to collect and classify. Unlike CPTs, BPTs are deliberately created effects that do not follow a fixed formula, cannot be repeated without losing their punch, and require a

92  •  C hapter 5

much more specialized environment. This is not to say that BPTs do not present common features—if they did not, they would be totally impermeable to narrative theory. Eventually, a theory of plot design will have to collect readers’ personal examples of BPTs and investigate the principles that produce these effects; but it is much easier to begin the theory from the weeds of the narrative flora than from the rare flowers whose sighting constitutes a memorable event, for the weeds, too, are narrative species from which we can learn something about plot design.

Cheap Plot Tricks The most productive (if I may say so) source of CPTs is the disregard of what Aristotle regards as the function of the poet, namely “not to say what has happened but to say the kind of thing that would happen, i.e. what is possible in accordance with probability and necessity” (5.5; 1996, 16). Most of my examples of CPTs involve coincidence, which, by definition, is a phenomenon of low probability because it is the product of an accidental intersection of two independent causal chains. The degree of probability of a coincidence is inversely proportional to the size of the pool of events that are possible in a certain situation; hence the “small world” effect created by coincidences: it is as if a large world had shrunk, leaving fewer possibilities. As Hilary Dannenberg has shown in her fascinating book Convergence and Divergence: Plotting Time and Space in Narrative Fiction (2008), narrative has never really outgrown the plot device of coincidence. If we decided to expurgate events of low probability in the name of artistic plotting, we would deprive ourselves of the main source of tellability, namely the report of unusual situations, and very few stories would survive. But our tolerance of extraordinary coincidence has decreased through the ages as the demand for realism has increased. Few of us are still fascinated by the highly contrived tales of shipwrecks and reunion of long-lost lovers that delighted readers in the Renaissance and Baroque periods. This evolution in taste explains why so many of my examples will come from the seventeenth century; but nowadays, CPTs are still widely found in popular culture, especially in film.

CPT 1: Extraordinary Coincidence: Being at the Right Place at the Right Time My first example of CPT, from Madame de Lafayette’s La Princesse de Clèves, ignited a memorable literary controversy. Shortly after publication, in 1678,

P lot  •  93

the periodical Le Mercure galant asked readers to express their opinion of this particular episode. The question generated a storm of responses anticipating the lively discussions of books, movies, and video games that one finds today on the internet. Mademoiselle de Chartres, a virtuous young beauty, arrives at the court of Henri II, King of France. The Prince de Clèves falls in love with her and asks for her hand. The princess, who has no experience of love, accepts, despite her lack of special feelings for him. After the wedding, she meets M. de Nemours, the most eligible bachelor at the court (though a notorious philanderer), and they fall instantly and passionately in love. But Mme de Clèves is determined to fight her passion, and she never gives evidence of her love to Nemours. To avoid temptation, she retreats to her country house. Nemours learns about her whereabouts and goes hunting in the area. He gets lost, and suddenly finds himself near a pavilion on the Clèves estate. Soon the princess and her husband walk by and sit on a bench right in front of him. From his hiding place, Nemours hears the princess confess to her husband that she is in love with another man, and that she needs to stay away from him. Though she never names the object of her love, the ecstatic Nemours under1 stands through circumstantial evidence that he is the one she has in mind.

The focus of the controversy was not, as one would expect, the presence of Nemours on the scene but the fact that Mme de Clèves confesses her love for another man to her husband, when nothing forces her to do so, since she is not guilty of any infidelity. According to the responses (Goldsmith 1998), most readers believed that no woman in her right mind would make such a confession. Today’s readers are much more understanding of Mme de Clèves’s action because it is well prepared within the text, both through the character of the heroine and through certain declarations of M. de Clèves that suggest his appreciation of honesty. But if my response is typical, modern readers are more troubled by the highly improbable spatial and temporal convergence of life paths that allows M. de Nemours to be present on the scene and to eavesdrop on Mme de Clèves’s confession. The artifice of the overheard confession allowed Mme de Lafayette to solve a thorny design problem. She wanted her heroine to be consumed by love, but she also wanted her to maintain the highest moral standards. These standards prevented the princess from giving any deliberate sign of love to M. de Nemours. On the other hand, the plot could not proceed to its tragic conclusion (and celebration of Mme de Clèves’s fortitude) without M. de Nemours

1. This summary and others that follow, except where noted, are mine.

94  •  C hapter 5

acquiring firm knowledge of the private feelings of Mme de Clèves. The CPT not only solves the problem of the transmission of information from a private to a public domain (for Nemours will gossip about it, creating a rumor that will eventually reach the unfortunate husband); it also kills two birds with one stone by awakening in M. de Clèves a jealousy that will make him die of a broken heart, through the literalized metaphor of another CPT. This death frees his widow to accept M. de Nemours’s marriage proposal; but for a secret reason variously interpreted as guilt, heroic self-control, resistance to social pressures, exceptional character, or fear of love, the princess turns down her suitor and chooses instead a life of penitence. The episode of the overheard confession blatantly serves the interests of the story and the goals of the author at the expense of verisimilitude. It was also in the name of verisimilitude that seventeenth-century readers criticized Mme de Clèves’s action, though they did not object to the presence of Nemours. The standards obviously differ: for the seventeenth-century reader (as for Aristotle), verisimilitude meant integrity of character and conformity with an idea of human nature, while for those modern readers taught by critical theory to distrust this notion, and who consequently cannot rely on models of “natural” behavior, the idea of verisimilitude is transposed from the realm of psychological motivations to the realm of the statistical probability of external events.

CPT 2: The Telltale (Lost and Found) Letter This device is found in another episode of La Princesse de Clèves: A gentleman of the court gives a letter to the reine dauphine, Marie Stuart, that supposedly fell from the pocket of M. de Nemours. The dauphine then gives it to Mme de Clèves. It is a love letter written by an unnamed woman, and it arouses feelings of jealousy in the princess. But it turns out that the letter was lost by another gentleman, who wants to retrieve it to avoid compromising his mistress. He writes a note to Nemours asking him to get it back from the dauphine. After the dauphine tells him what she did with the letter, Nemours visits Mme de Clèves and gossips about the note, to the great relief of the princess, who now understands that her jealousy was unfounded.

Plots are heavily dependent on the circulation of information, whether successful, inadvertent, or blocked. Lost and found private letters constitute an overly convenient way to make information fall into the wrong hands (or

P lot  •  95

minds). In this passage the author faced the problem of making the princess experience jealousy—a proof of her love—while preserving the “innocence” of M. de Nemours, who, since the day he met the princess, has completely given up his philandering habits. After giving the princess grounds for jealousy, the author had to clear Nemours in her eyes. Mme de Lafayette does so through a convoluted scheme that invalidates the information inferred by the princess: the letter, after all, did not fall from the pocket of M. de Nemours, and the man who gave it to the queen was mistaken about its origin, though we never learn why. This scheme illustrates another common plot device: making characters act or think on the basis of false information and gratuitously withdrawing this information once the reaction of the character has taken place because it is not compatible with the planned development of the plot. In the case of La Princesse de Clèves, the purpose of the false information is to allow a strictly private event of self-understanding, but in my next example it has much more open, and damaging, consequences.

CPT 3: The False News From Racine’s tragedy Phèdre: Phèdre, the second wife of Theseus, is madly in love with her stepson Hippolyte. Ashamed of feelings considered incestuous, she decides to die. Just in time, however, a messenger arrives with the news of the death of Theseus. In the ensuing scene, Phèdre reveals to the horrified Hippolyte her consuming passion for him and asks him to kill her. His refusal to do so gives Phèdre some desperate hope. But these hopes are shattered when news arrive that Theseus, far from being dead, will soon return home (so soon, indeed, that he will be on stage in the next act).

Here again the CPT allows the author to resolve a contradiction. Phèdre may have “incestuous” feelings (or what passed for such in the seventeenth century), but she has an acute sense of her honor and would not commit adultery. Racine needs Theseus to be dead for Phèdre to declare her love to Hippolyte, and he needs him to be alive for Phèdre to be guilty of improper conduct. By operating a complete reversal of situation, the CPT of the false news coming out of the blue allows Racine to take advantage of two mutually incompatible situations without violating the laws of nature by making Theseus return from the dead, and without turning Theseus into a human version of Schrödinger’s cat: dead and alive at the same time.

96  •  C hapter 5

CPT 4: The Calumniator Credited The resolution of Phèdre involves a classic CPT known as the convention of the calumniator credited. When the news of the imminent return of her husband reaches Phèdre, she is afraid that Hippolyte will denounce her. To avoid punishment, she lets her confidante, Oenone, falsely accuse Hippolyte of sullying her honor. Furious with his son, Theseus banishes him from Thèbes. Overcome by guilt, Phèdre decides to clear Hippolyte’s reputation by telling the truth to Theseus, but she changes her mind when she learns that Hippolyte is in love with another woman. In the end, Hippolyte is killed by a sea monster, and Phèdre commits suicide.

Why did Theseus believe Oenone without giving Hippolyte a chance to defend himself? The convention of the calumniator credited, named in 1934 by the Shakespeare scholar Elmer Stoll, asks the audience to bracket out this question. Martin Steinmann defines the convention as follows: “If X, Y, and Z are characters in Elizabethan drama [or French classical, as my example shows], if X calumniates Y to Z, and Z, without proof or serious investigation, credits (that is, believes) X’s calumnies, then we may not interpret Z’s crediting them as a . . . natural sign” (1981, 288). If the act of crediting were interpreted as a natural sign, we would attribute it to a mental feature of the character. We would, in other words, infer that Theseus (or Othello) is either naïve or stupid, judging him by the same criteria we would use with a real-world person who behaved in this way. Instead we suspend judgment, and we assume that Theseus is made to act in the interest of the plot. As Steinmann writes, “The convention serves the specific purpose of providing interesting complications. It is the usual trade-off: we accept lack of realism for an artistic reward” (255). The convention of the calumniator credited lies on the borderline between CPTs and PHs (a category to which I return shortly). Like all CPTs, the convention consists of an easily recognizable and frequently used type of event that solves a standard narrative problem (in this case, making characters take actions that are not quite in line with their moral features), but in doing so it creates a plot hole because it opens a gap in the psychological motivation of the characters’ actions.2

2. Roland Barthes was fully aware of this dilemma between psychological verisimilitude and the demands of plot. See my discussion of his treatment of “Sarrasine” in chapter 4.

P lot  •  97

CPT 5: Deus Ex Machina and Irrational Events All the CPTs discussed so far have a preparatory function: they complicate a situation, and they steer the plot toward a climax. A standard metaphor describes this process as tying the plot into a knot (French: nouer l’intrigue). The knot must be eventually disentangled for the plot to reach a proper conclusion; but, as Aristotle observes, “Many poets are good at complication but handle the resolution badly” (8.6; 1996, 30). A facile way to conclude a story, when the author has painted the characters into a corner by producing a situation that cannot be resolved by natural means, is the classic device of the deus ex machina. The term refers to the habit in ancient Greek drama of lowering a god with a crane onto the stage. Aristotle’s objection to the device stems from its irrational and arbitrary character: “Clearly . . . the resolution of plots should also come from the plot itself, and not by means of a theatrical device, as in the Medea, or the events concerned with the launching of the ship in the Iliad” (8.1, 25). In the Medea, the heroine is saved by a chariot from heaven after she murders her children; in the Iliad, it takes the intervention of the goddess Athena to force the Greeks to resume fighting, after they have taken at face value Agamemnon’s ironic suggestion that they abandon the siege of Troy. Nowadays the term deus ex machina is extended to any unexpected event that brings a happy ending from the outside when the characters have 3 exhausted all possibilities of improving their own fate. The god could be replaced by a messenger from the king who saves the hero and his family from being evicted from their house, after the hero has given all his property and promised the hand of his daughter to a swindler who poses as a deeply religious man (Molière, Tartuffe); by an attack of microbes that saves the earth from Martian invaders (H.  G. Wells, War of the Worlds); by a timely solar eclipse that saves the comic book hero Tintin from being sacrificed to the sun god (Hergé, Prisoners of the Sun); or by a bird that distracts the attention of the villain as he is holding the heroes in the cross-hairs of his gun (Dan Brown, The Da Vinci Code). In the better cases, the deus ex machina effect can be justified by thematic considerations: for instance, Molière’s use of the device, by stressing the artificiality of the ending, could be interpreted as a way to tell the spectator that all ends well in this play because it is a comedy, and comedies are supposed to make you laugh; but the Tartuffes of this world would succeed, because in real life, hypocrisy reigns supreme and controls 3. It is worth noting that the use of the expression deus ex machina implies a benevolent god: there is no such thing as a diabolus ex machina, even though it is quite possible for a random event to cause a tragic ending (cf. Nabokov’s Pale Fire, where the bullet of an assassin kills the wrong person).

98  •  C hapter 5

society. In the worst cases, the device is nothing more than a convenient way to satisfy the reader’s need to see the hero suffer and then triumph. In these cases, the particular nature of the rescuing action, or the personality of the Deus character do not really matter. The avoidance of the deus ex machina effect is a particularly acute problem in fantastic narratives that rely on magic. When used without restraint, magic is the ultimate CPT, since it can take characters out of any situation, dispensing the author from constructing materially and psychologically credible solutions. If characters can solve any problem through magic spells, how can they or their protégés ever fall into serious trouble, and how can the plot create conflict, the primary source of narrative interest? Alternatively, if a character has supernatural powers and fails to use them to get out of trouble, how can this be justified? Authors generally avoid the two pitfalls of overuse and unexplained nonuse of the supernatural by treating magic as a resource that exists in limited supply (as it does in video games). As Patrick Colm Hogan observes (2003, 219), even in the most fantastic worlds, narrative uses the supernatural very sparingly. Magic is not a force that allows its masters to do anything they want but a specialized weapon whose efficiency depends on an understanding of its proper handling. Its use may be limited to certain places and times, and it must respect certain rules which may override, but never totally cancel, the laws of nature. Most importantly, magical abilities are distributed between many characters, and they exist in many varieties, so that every effective use of magic must take into consideration the supernatural resources of the antagonist. For instance, when Harry Potter fights Voldemort, he must be able to disable his enemy’s spells through his own magic tricks. Thanks to this gamelike, rule-governed character, magic remains compatible with the rationality that Aristotle regards as an essential condition of artistic plotting.

CPT 6: The Interrupted Action An alternative to the deus ex machina ending that shares with it a random character is the sudden, logically unmotivated interruption of action. An example of plot that just stops is the ending of the wildly popular TV series The Sopranos on June 10, 2007. The long-awaited last episode concludes not with a spectacular event that brings the “mother of all closures,” as would the death of the main character, Tony Soprano, but with an ordinary family meal at a nondescript diner. The screen turns black just as the Soprano daughter, Meadow, arrives late at the gathering after having trouble parking her car, and

P lot  •  99

after a delay the credits appear. Some spectators were so surprised to see the episode end on this scene that they believed they had lost their TV signal. Opinions were divided as to whether this ending represents a brilliant comment on the lack of closure of life, and consequently on the disparity between narrative form and reality, or a cheap cop-out demonstrating the writer’s inability to tie the strands of the plot in a satisfactory way.4 One thing is, however, certain: the lack of closure left the door open to a revival of the series.5 Should we regard an interrupted ending as a CPT? Those who object to the device can point to the fact that it can be used to terminate any plot, regardless of the particular circumstances. It does not grow bottom-up from the events that make the narrative world evolve but is applied top-down, as a seemingly arbitrary decision of the creator. Those who find the ending of The Sopranos a stroke of genius might reply that no plot twist is inherently objectionable, and that the sudden interruption of action is redeemed by its existential significance. But what ideas can an author express through this device, besides the rather predictable message that “life goes on”? This overly general lesson does not, however, prevent readers and spectators from finding more specific meanings in the particular details of the scene that interrupts a narrative. In the case of The Sopranos, the internet was full of speculations about what the author was trying to say by having Tony play a particular song on the jukebox or by making a man wearing a “Members Only” jacket walk past him on his way to the bathroom, and about whether the prolonged darkness that follows the last frame means that Tony has died. Interrupted endings manage to be at the same time the refusal of the inherent trickery of plot and the ultimate CPT.

Plot Holes If you have ever had the annoying experience of watching a movie and not quite understanding why the characters are acting the way they are, you may have missed crucial information, something easy to do with a streaming medium that controls the pace of its display and that distributes attention among many types of signs. Alternatively, you may have noticed a plot hole. While in CPTs authors play god to ensure a proper narrative arc, in 4. It is interesting that in narrative matters the metaphor of the knot is used in two opposite ways: in one interpretation, conflict creates a knot and resolution unties it; in the other, plot is compared to a rug and the ending must tie the loose ends together, to prevent unraveling. 5. According to a recent interview with the writer, David Chase, Tony Soprano dies at the end of the episode (Feinberg 2021). This, however, was not clear to many spectators.

100  •  C hapter 5

plot holes they ignore, or fail to notice, logical inconsistencies that should normally prevent the progression of the plot toward its outcome. From the reader’s point of view, PHs are much more disturbing than CPTs because the latter are immediately recognizable whereas the former arouse the suspicion “Am I stupid? Have I missed something?” It is a relief for readers to be able to attribute the inconsistency to the author’s ineptitude rather than to their own mental deficiencies. In the common use of the term, plot hole designates an inadvertent inconsistency in the logical and motivational texture of a story. This situation must be distinguished from ontological paradoxes that drill holes through the intelligibility of a fictional world. Certain narratives, such as Kafka’s Metamorphosis or Emmanuel Carrère’s La Moustache (a story discussed in chapter 8 in which the past of the hero is constantly changing), are built around fantastic events that defy understanding, even though their world is consistent in all other respects. While realistic narratives construe the fictional world as an apple, this is to say, as a fully filled sphere of rational events, and narratives with plot holes construe it as a wormy apple into which readers may bite without noticing the worms, narratives with ontological paradoxes construe a world that looks like Swiss cheese, with zones of irrationality clearly delimited from areas accessible to logical inferences, so that they do not throw the entire fictional world into cognitive chaos. (See chapter 8 for more on the Swisscheese interpretive strategy.) Inadvertent plot holes are particularly frequent in film because the emphasis of the medium on visible action, its time constraints, and its allegiance to highly dramatic effects require a tightly plotted storyline. The more action one squeezes into a limited temporal frame, the greater the need for logical (i.e., causal or motivational) connections, but also the greater the probability that some of these connections will be overlooked by the scriptwriter. It is also easier to get away with plot holes in film than in print narrative because of the streaming nature of the former. The spectator of cinema is much more focused on the present than is the reader of a novel, who can interrupt reading to think about past events or reread earlier passages. If movie spectators give too much thought to the plot holes of previous scenes, they will be unable to keep up with the current developments. The classic movie Citizen Kane offers a relatively harmless example of plot hole: In Orson Welles’ Citizen Kane, a group of reporters is trying to discover the meaning of Kane’s dying words “Rosebud.” However, Kane dies alone. When Welles was informed of this, he reportedly stared for a long time before saying, “Don’t you ever tell anyone of this.” (Wikipedia, “Plot Holes”)

P lot  •  101

In Citizen Kane, the plot hole is a minor oversight which could have been easily fixed, if the scriptwriter had been aware of it. For instance, Kane’s last word could have been heard by a servant, and he could have died alone several hours after uttering it. Moreover, of the two incompatible events that create the hole, only one—uttering “Rosebud” as a final word—implicates the future development of the story; the fact that Kane died alone, a comment on his life and character, has a symbolic rather than a causal function, and it would be easier to delete without damaging the logical integrity of the plot. The truly unbridgeable plot holes involve strategic decisions from the characters that blatantly violate common sense. While characters may not always act as we would if we were placed in the same situations, we expect of them a minimum of rationality,6 unless, of course, they suffer from impaired mental abilities. Consider the character’s motivation in two examples. A mother tells her daughter, Little Red Riding Hood, to go through the forest and to bring some food to her ailing grandmother. She warns the little girl not to talk to strangers. On her way, Little Red Riding Hood meets a hungry wolf and tells him about her mission. The wolf runs to the grandmother’s house, eats her, and takes her place in bed. When Little Red Riding Hood arrives she mistakes the wolf for the grandmother. After a conversation during which he pretends to be the grandmother, the wolf jumps out of the bed and eats Little Red Riding Hood.

Question: Why didn’t the wolf eat the little girl on their first encounter, if he was really hungry? By delaying his repast, wasn’t he running the risk of losing her to another wolf? Answer: he does not eat Little Red Riding Hood on the 7 spot because it makes a better story. As I have argued in Ryan (1991; chapter 11), most narratives are created not prospectively but retrospectively. Rather than charting the evolution of an initial situation, the storyteller imagines a climactic scene, a situation of high tellability, and constructs a causal chain of events that leads to the target situation. Here the target was the highly dramatic, slightly comic, and visually appealing scene of the encounter of the heroine with the wolf disguised as the grandmother. Another example of defective motivation comes from Dan Brown’s megabestseller The Da Vinci Code: 6. Cf. Schaeffer and Vultur’s (2005) definition of mimeticism, discussed in the introduction. 7. The anonymous reader of the manuscript informs me that in Charles Perrault’s version, the wolf does not eat Little Red Riding Hood during the first encounter because woodcutters are nearby. This information certainly closes the logical gap. But it is lost in many of the countless retellings.

102  •  C hapter 5

The curator of the Louvre is murdered in the museum. Before he dies he draws a strange symbol with his blood, together with the message “Get Robert Langdon.” The police summon Langdon, a professor of symbology, to the scene, where he is joined by Sophie, a cryptologist and the granddaughter of the victim. Langdon soon realizes that the police suspect him of the murder and want to arrest him. This sets in motion a hectic manhunt during which Sophie and Langdon escape the police through a series of daring moves, as if their lives depended on it, while attempting at the same time to decipher a chain of riddles devised by the dead man. In the end, however, the very same cop who initiated the chase arrives like a deus ex machina to arrest the real murderer, who had just been disarmed by Langdon and Sophie.

In the course of their attempt to escape from the French police, Langdon and Sophie perform many daring actions that put their lives at risk: using a Da Vinci painting as shield in a move to disarm the cops, driving a jeep at high speed through rough terrain with the cops breathing down their necks, stealing an airplane and flying away just as their pursuers are catching up to them. But since there is no solid evidence against them, what did Langdon and Sophie have to fear from being arrested besides being interrogated and released? Moreover, by attempting to flee, aren’t they incriminating themselves? The author is aware of this inconsistency, and builds rather unconvincing arguments to justify the behavior of both the police and the heroes: the cop is motivated by religious faith and a need for promotion; Langdon flees in the hope of taking refuge at the US embassy, where he believes he will be protected by his government. The novel treats French police as if France were a totalitarian state where innocent people can be arbitrarily arrested, summarily tried, and sent to rot in jail, rather than a modern democracy that presumes people to be innocent until found guilty. But a more rational behavior on the part of Langdon and Sophie (or on the part of the police) would have deprived readers of a breathtaking chase that maximizes dramatic effect by giving the heroes the double task of deciphering a chain of riddles while trying to escape from their pursuers.

Evaluating Plot Devices Literary taste is historically variable, and, as the example from La Princesse de Clèves demonstrates, so are judgments of what constitutes a CPT. It is always dangerous for critics to interpret their own reactions as typical of a wider reading public, but here I assume indeed that my evaluation of the devices

P lot  •  103

described above as cheap is reasonably shared among at least a certain class of readers—let’s call them “academic.” This does not necessarily mean that we dislike the stories that make use of these devices; on the contrary, we may enjoy them cum grano salis, as an example of literary kitsch, or we may decide that the narrative situation to which they lead was well worth the cost of a CPT. Our aesthetic evaluation of plot devices can be captured by a simple economic principle: if a device defies our willingness to suspend disbelief, was the reward worth the sacrifice, or could the reward have been obtained at a cheaper cost, or at no cost at all? In other words, can we think of a better solution to the plotting problem, one that gives the impression of growing from within the narrative situation rather than depending on the intervention of too freely created external events? The idea of a trade-off raises the question of what is to be gained and what is to be lost by using CPTs. The advantages are obvious. All my examples of CPTs and PHs allowed the story to reach a situation of intense dramatic tension. In La Princesse de Clèves, the CPT of the extraordinary convergence of life paths leads to the highly emotional situation of Nemours realizing that he is loved; in Phèdre, false news incites the heroine to compromise her reputation; and in deus ex machina plots, the CPT satisfies the audience’s desire for a happy ending. Now that the emotional impact of narrative has regained favor among literary critics (after an eclipse under New Criticism and deconstruction), it would take a Cartesian curmudgeon to dismiss without trial the violations of narrative logic and the conventional devices that lead to heart-rending scenes. Besides preparing situations of great emotional impact, CPTs also steer the plot toward a satisfactory climax and resolution, whereas plot holes allow the narrative to jump over potential logical obstacles. If they are so beneficial, why do we object to them? The rejection of CPTs by some audiences is symptomatic of a distrust of plot as an adequate way to represent reality. As historian Hayden White argues, stories imposes on reality a form that is alien to it: Does the world really present itself to perception in the form of well-made stories, with central subjects, proper beginnings, middles, and ends, and a coherence that permits us to see the end in every beginning? Or does it present itself more in the form that the annals and chronicles suggest [i.e., as a list of events that leaves out causal relations]? (1981, 23)

White’s implicit suggestion that “how the world presents itself to perception” constitutes the goal of historical representation and that annals and chronicles—notoriously plotless forms of narrative—are more true to life

104  •  C hapter 5

than emplotted stories, rests on a questionable belief in a “raw” perception untainted by narrative interpretation as well as on an elimination of mental phenomena from reality. If history is made of events, and if most events are the result of actions performed by thinking human beings, historiography must take minds into account, and the content of these minds must be inferred, since it is does not present itself to perception. (One of the major concerns of cognitive psychology, known as “theory of mind” [Zunshine 2006] is indeed our ability to construct other people’s thoughts.) Still, White’s distrust of emplotment as an adequate means of representation has been very influential for the practice of historiography. Though his critique does not target fiction—his point, rather, is that narrative history is no more truthful than fiction because it imitates literary forms such as tragedy, comedy, and farce—its relevance extends to any literary work whose aesthetic and ethical goal is to present an image of what could happen in the real world. Plot, in this perspective, is a form imposed top-down on reality rather than growing out of it. As stereotyped devices borrowed from literary tradition, devices that have traveled, virtually unchanged, through countless fictional worlds, CPTs are the worst culprit and the most blatant evidence of the artificiality of plot. The reader’s acceptance of CPTs hinges on many factors. As the example of La Princesse de Clèves demonstrates, one of them is the historical variability of literary taste. After flourishing in the Renaissance and Baroque age, and well into the eighteenth century (think of the plot of Beaumarchais’s comedy The Marriage of Figaro and of the Mozart opera it inspired), CPTs saw a sharp decline in the nineteenth century. Though Dannenberg (2008) observes some examples of amazing coincidence—the main source of CPTs—in Austen, Brontë, and Dickens, these situations, which involve characters discovering that they are related or have common acquaintances, appear much less contrived than the CPTs of the Baroque novel because they take place in a much more restricted social and geographic environment. For instance, given the closed world of the British gentry in which Jane Austen’s Pride and Prejudice takes place, the odds against Elizabeth Bennet’s cousin Collins being the rector of the parish presided over by Darcy’s aunt Lady Catherine are infinitely lower than the odds against Nemours eavesdropping on Mme de Clèves’s confession. Small-world effects are phenomena that do occasionally occur in the real world, and when they do, we feel an irresistible urge to tell stories about them. Their presence in the novels of realism therefore goes with the territory of social reality. CPTs continued their decline in modernism, together with heavily plotted stories. But as Dannenberg observes, they are presently enjoying a minor revival because their contrived and conventional nature can be used in support of the postmodernist/structuralist view that language constructs, rather

P lot  •  105

than reflects, reality, and that thought is conditioned by an arbitrarily configured system of signs. The use of CPTs in late twentieth-century literature is no longer a “naïve” attempt to immerse the reader in a fictional world that takes temporarily the place of reality but a self-reflexive, or metafictional, device that underscores the textual origin of the fictional world. For the postmodernist, no plot twist is so cheap that it cannot be redeemed by irony. As Julian Barnes writes: And as for coincidence in books—there’s something cheap and sentimental about the device . . . the sudden but convenient Dickensian benefactors; the neat shipwreck on a foreign shore which reunites siblings and lovers.  .  .  . One way of legitimizing coincidences, of course, is to call them ironies . . . I wonder if the wittiest, most resonant irony isn’t just a well-brushed, welleducated coincidence. (1985, 67)

Twentieth-century examples of ironic, parodistic, and self-reflexive uses of CPT abound: for instance, the rescue in extremis of the heroes of Brecht’s Three Penny Opera through the deus ex machina arrival of the king’s mounted messenger (an irony also present in Brecht’s eighteenth-century intertext, John Gay’s The Beggar’s Opera);8 or the discovery by one of the two protagonists of A. S. Byatt’s Possession that she is the descendant of the two Victorian poets whose love affair they have been studying—an obvious parody of romantic plots. The ironization of CPTs is so typical of the postmodernist spirit that it threatens to become yet another CPT—or rather a meta-CPT. This could explain why the self-reflexive stance has now percolated from “high” literature to popular culture—which, according to Steven Johnson (2005), is becoming more and more sophisticated as people become more literate in its media of dissemination: film, TV, and video games. In this example, from the Wikipedia entry on plot holes, a literalized metaphor is used as a means of teletransportation: In Tiny Toon Adventures: How I spent my Vacation, a plot hole (portrayed by a physical hole) is used to transport Babs, Buster and Byron back to Acme University. Babs makes the comment “A plot hole. I wondered how the hack writers would get out of this one.”

8. At the end of the play, the character of the Beggar declares that Macheath deserves to be hanged, but since this is a comedy, and since comedy does not tolerate tragic endings, he will be saved from the gallows and reunited with his wife, Polly (actually, one of his wives). The case of The Beggar’s Opera provides an excellent example of the metafictional tendencies of eighteenth-century literature and of its affinities with postmodernism.

106  •  C hapter 5

If ironic self-reflexivity can allow contemporary authors to enjoy the benefits of CPTs without paying the price—that is, without being themselves accused of bad plotting—it should work just as well to protect the reputation of authors of the past. In a move typical of deconstruction, the dominant school of literary criticism at the time of her article’s publication, Dalia Judovitz exonerates Mme de Lafayette’s use of CPTs by reading her use of verisimilitude-transgressing events as an allegory of what she regards as the unavoidably “fictional” nature of representation: The novel’s identification of representation and fiction announces the emergence of the aesthetics of implausibility, for art now becomes the expression of the true character of representation. This new aesthetics is no longer defined by its social or ethical adequacy, by its adherence to some ultimate truth but by its representational character, that is to say its own truth as selfimposed, willed and created representation. (1984, 1054–55)

In such a reading, the novel’s literary value resides not in its presentation of the ethical dilemma of Mme de Clèves but in its awareness of the allegedly “true” nature of representation, which paradoxically resides in a decoupling from truth. The possibility of truth is thus shifted from the representational to the metarepresentational level. A second factor of acceptability is genre. Historical periods are marked by the predominance of different types of fictional worlds. These world types tend to develop into culturally recognized genres, such as pastoral romance, chivalric novels, and comedies of errors for the sixteenth and seventeenth centuries, or the fantastic, science fiction, detective stories, and historical novels for contemporary literature. The more realistic a genre—the closer its world to our model of everyday reality—the less tolerant readers will be to the use of plot twists that stretch their willingness to suspend disbelief. CPTs and PHs are much more acceptable in gothic novels, horror stories, science fiction, medieval fantasy, and magical realism than in historical novels, psychological novels, postcolonial novels, and detective stories (which must maintain a certain level of cleverness). The impact of genre on the reader’s evaluation of plot is illustrated once again by La Princesse de Clèves. The readers of the seventeenth century judged Mme de Lafayette’s novel according to a Baroque aesthetics that cherished convoluted plots, stories within stories, extraordinary coincidence, and exotic settings. Not only did Mme de Lafayette contribute to the tradition of the seventeenth-century romance in her earlier novel Zaïde, a complicated tale of romantic love, adventure, kidnapping, and journeys to faraway countries with numerous levels of embedding; she demonstrates the

P lot  •  107

indebtedness of La Princesse de Clèves to Baroque aesthetics through several framed tales that digress from the main storyline. But, for the modern reader, La Princesse de Clèves is the first genuinely historical novel of French and maybe of European literature. The episode of Nemours overhearing Mme de Clèves’s confession is much more shocking in a work that features mostly historical characters, relates historical events (such as the death of King Henri II in a duel), and relies heavily on documents and testimonies of court life in the sixteenth century than it would be in the pastoral romances and chivalric novels that flourished at the time.9 Our acceptance of CPTs is also affected by their location within the narrative arc. A preparatory CPT is partially forgotten when the story reaches its climax, but a deus ex machina ending lingers in the reader’s memory. A story-ending CPT represents a failure to find a satisfactory resolution to the narrative conflict; a preparatory CPT at least holds the promise of future satisfaction. Another reason we are more tolerant of conflict-creating than of conflict-resolving CPTs is that we want the characters to (appear to) be autonomous agents who exercise some degree of control over their own lives rather than the puppets of authorial whimsy. Aristotle sensed the importance of making characters masters of their own fate when he recommended limiting interventions of the gods to the prehistory or posthistory of the events shown on the stage: “There should be nothing irrational in the events themselves, or failing that, it should be outside the play, as for example in Sophocles’s Oedipus” (8.1; 1996, 25). Even when they occur early in the story, CPTs bother us through their excess of tellability. By this I mean that if the events that we regard as CPTs happened in real life, they would be interesting enough to form the point of a story, but they fail to convince in a fiction because it is just too easy for the author to make them up. Aristotle viewed the task of the poet as more deeply philosophical than the task of the historian because the poet must convince the reader of the plausibility of the narrated events whereas the historian reports events whose possibility cannot be contested, since they actually happened (Poetics 5.5; 1996, 16). (Let’s remember that the actual is a subset of the possible.) Our ambiguous attitude toward CPTs and PHs reflects the paradoxical nature of the expectations we bring to narrative: on one hand, we want some degree of verisimilitude; otherwise, we could never relate emotionally to the characters nor follow their reasoning. On the other, we demand the tellability of extraordinary events. We want our stories to be true to life, 9. The indebtedness of La Princesse de Clèves to the plots of pastoral romance has been noted by Thomas Pavel (2003, 128–29).

108  •  C hapter 5

in that they should record the efforts of humans to adjust to the genuine randomness of destiny, but we also want them to display the purposefulness of narrative form, which denies randomness. The obviously calculated pseudo-­ randomness of CPTs and the inadvertent inconsistencies of PHs guarantee narrative form and tellability at the expense of credibility. But while the pursuit of narrative excitement at all costs leads to a dependency on CPTs, the caveat of the opposite strategy is to fall into an aesthetics of triviality that views life as basically repetitive and boring, and that associates literary value with the representation of small and ordinary events.10 An important aspect of narrative art lies in finding the right balance between the conflicting demands of verisimilitude and tellability. From a literary point of view, the most significant criterion of acceptability for a plotting device is its thematic adequacy and symbolic value. We are not bothered by the highly incredible sequence of catastrophes and miraculous rescues of the heroes in Voltaire’s Candide, because the point of the story is to challenge Leibniz’s pronouncement that “everything is for the best in the best of possible worlds.” Exaggeration is perfectly in tune with the satirical mood of the story. The role of the miraculous rescues is not to demonstrate the workings of Providence but, more plot-functionally, to keep the heroes alive and to allow the author to pile up more catastrophes in their life path. An even more striking example of a thematically grounded transgression of probability comes from Sophocles’s Oedipus Rex. Here the plotting of the author stands for the will of the gods who hold the strings of human lives, and his heavy hand in arranging the circumstances that make Oedipus guilty of parricide and incest can be interpreted as an allegory of the inescapability of fate. In La Princesse de Clèves, by contrast, the heroine’s refusal to give in to her passion signifies the possibility of remaining in control of one’s life despite the accidents of fate, and the author’s recourse to extraordinary coincidence clashes with this moral.

Conclusion The point of my investigation is not to castigate the use of CPTs but to gain a better understanding of the mechanics of plot. Narrative is said to consist of story and discourse, but the vast majority of narratological work has focused either on the latter, or, with sociolinguistic approaches, on the pragmatics of 10. My Struggle, the 3,600-page autobiographical novel of Karl-Ove Knausgaard (2013–19), which reports in minute details his day-to-day activities, is a striking example of this aesthetics of triviality.

P lot  •  109

narrative communication, leaving “storyology”—the study of the logic that binds events into plots—mostly to scriptwriters and authors of how-to manuals. The present chapter looks at the opposite of good stories, but as the negative images of BPTs, CPTs and PHs teach us indirectly about efficient plot construction. By asking of CPTs what design problems they are supposed to solve, I hope to have sketched an approach to plot that aims at the heart of narrative logic. CPTs, BPTs, and ordinary PTs are cogs and wheels in a machine engineered to produce certain effects on the user, and their understanding begins with a fundamental question that was first formulated by Vladimir Propp: what is their function for the story as a whole?

CHAPTER 6

Mimesis and Diegesis Complementing Each Other

The primary goal of a cognitive approach to the reception of narrative fiction should be to describe how users experience the text, how they relate to what is presented to them, and how they imagine it, as opposed to what they know about it. In this chapter, I propose to address the question of the user’s act of imagination with the help of Plato. In the third book of the Republic, Plato introduces a model of narrative based on the notions of diegesis and mimesis, or telling and showing. He does so as an argument for excluding poets from the Republic, for he thinks it immoral for poets to imitate bad people, but the distinction has been considered a cornerstone of poetics for centuries, and it has been reinterpreted in many different ways. While mimesis and diegesis are often viewed as binary opposites, Plato used the term diegesis to mean narrative in general, and he distinguished three modes of narration, this is to say, three species of diegesis: a narrative can be either simple [haple diegesis in Greek], or produced by imitation [diegesis dia mimeseos], or by both together [diegesis di amphoteron] (Plato 1968, 71; Halliwell 2014). By simple narrative, Plato means that “the poet himself speaks and does not attempt to turn our thoughts elsewhere,” while by mimesis, “he speaks as though he himself were [a character] and tries as hard as he can to make it seem to us that it is not [himself] speaking” (71). How do these two features yield three modes of narration? If they are mutually incompatible, they should produce two types—the poet either speaks for 110

M imesis and D iegesis  •  111

himself, or imitates a character. On the other hand, if they can be combined, we should have four categories, corresponding to the binary numbers 00, 01, 10, and 11. What happens here is that the basic features can be combined, but the category 00 is not represented, since it would mean that the poet is neither speaking in his own name, nor imitating characters. This is how two times two becomes three rather than four. “Of poetry and tale-telling, one kind proceeds wholly by imitation—as you say, tragedy and comedy; another, by the poet’s own report—this, of course, you would find especially in dithyrambs, and still another by both—this is found in epic poetry and many other places too, if you understand me” (1968, 72). The many other places could include nonfictional and nonpoetic narrative genres such as conversational storytelling, and, in today’s culture, journalism or autobiography. The mimetic element in these mixed modes is represented by dialogue, where, indeed, the narrator speaks as the characters. But it would be difficult to find narrative genres that explicitly prohibit dialogue; therefore, Plato’s example of a pure representative of “simple diegesis” is the nonnarrative genre of the dithyramb, which later developed into lyric poetry, a genre of at best optional and certainly reduced narrativity. Moreover, when Plato mentions comedy and tragedy as examples of a purely mimetic narrative mode, he refers to what we would today regard as a distinct narrative medium, rather than as a narrative genre. Yes, dramatic texts are often considered a literary genre, but only when they are divorced from the performance through which they are meant to be actualized.

Comparing Conceptions of Mimesis and Diegesis It is interesting to note that while Plato regards all forms of poetry as species of diegesis, Aristotle builds his Poetics on the notion of mimesis, which translates as imitation. According to Halliwell (2014, 132), mimesis can be used in a narrow and in a broad sense. In the broad sense, it means representation, depiction, or expression in various media, while in the narrow sense it refers to dramatic enactment. Plato uses mimesis in the narrow sense, while Aristotle uses it in the broad sense to cover all artistic media: “Epic poetry and the composition of tragedy, as well as comedy and the arts of dithyrambic poetry and (for the most part) of music for pipe and lyre, are all (taken together) imitations” (Poetics 2; 1996, 3). The various arts can be differentiated from each other though their medium (i.e., image, language, music), through their object (what kind of thing or people they imitate), and through their mode, which corresponds to Plato’s forms of diegesis: “It is possible to imitate the same object in the same medium sometimes by narrating (either using a different

112  •  C hapter 6

persona, as in Homer’s poetry, or as the same person without variations), or else with all the imitators as agents and engaging in activity” (2.3; 1996, 5). Since Aristotle uses a term other than diegesis for narrating (apangellein, to report, according to Halliwell [2014, 133]), he recategorizes Plato’s forms of diegesis as different modes of mimesis, using this term to mean representation. The broader use of mimesis has persisted to the present; for instance, the subtitle of Kendall Walton’s classic book Mimesis as Make-Believe: On the Foundations of the Representational Arts (1990) suggests equivalence between mimesis and representation; similarly, when advocates of unnatural narratology describe certain kinds of narrative as antimimetic (Richardson 2006, 2016; Alber 2016), they do not equate mimesis with dramatic enactment: they rather mean that these narratives reject representation or do not represent events that could happen in the real world. Paul Ricoeur’s three kinds of mimesis (I, II, and III) are also more indebted to representation than to theatrical enactment. Diegesis has undergone a similar reinterpretation since its use by Plato. Even though film is a fundamentally mimetic medium, as are all image-based media, the terms diegetic and extradiegetic are used in film studies to indicate whether or not elements such as music belong to the storyworld presented by the film. The cinematic use of diegesis inspired Genette, though he associates the term with story rather than with storyworld, a concept he does not use:1 “With the same meaning [story], I will also use the term diegesis, which comes to us from the theoreticians of cinematographic narrative” (1980, 27n2). I find, however, his use of diegesis ambiguous: when he calls narrators intraor extradiegetic, this means that they are either internal or external to the storyworld, but when he distinguishes homo- from heterodiegetic narrators, this means that they do or do not tell stories about themselves. The first case aligns with film theory, but the second, by foregrounding the act of narration rather than the storyworld, is consistent with Plato’s use.

Semiotic Basis of Mimesis and Diegesis Here I propose to return to the narrow, Platonian sense of mimesis, by stressing its imitative nature, and by regarding imitation as a subspecies of representation rather than as synonymous with it. To imitate means to copy or simulate, and imitation rests on an iconic resemblance to the thing being imitated. According to the OED, an imitation is (among other definitions) “an act of imitating a person’s speech or mannerisms.” This is exactly what happens

1. I regard story as the temporal dimension of storyworld.

M imesis and D iegesis  •  113

in dramatic impersonation by actors, or in narrative dialogue. Plato’s simple diegesis (which I call henceforth narrative diegesis) does not involve imitation, because it relies on language, and the signs of language do not resemble that which they signify. They work through symbolic signification, not through iconicity, except in the rare case of onomatopes.2 The only thing that language can truly imitate, as Genette observed, is language itself: “The truth is that mimesis in words can only be mimesis of words” (1980, 164). Language can imitate language on the macro level, when a fictional narrative mimics a genre of nonfiction (for instance, a confession or a memoir); but more frequently, it imitates on the micro level, by having the narrative discourse quote the speech of characters. In dramatic acting, the reproduction of speech is part of a global act of imitation of a character’s behavior that also includes an imitation of gestures and actions. The imitation thus involves both visual and verbal components. In the dialogues of language-based narrative, by contrast, the imitation of speech is embedded in a storytelling act that represents through convention-based signification rather than through iconic resemblance. The model I am proposing regards representation as any use of signs that can evoke the thought of objects or events in the mind of the observer. Narrative is a subset of representation; it consists of a group of signs deliberately arranged by an author or authorial team (not necessarily by a narrator) to evoke in the audience a mental representation that fulfills certain conditions— the conditions that define stories. Narrative signs can belong to all three of Peirce’s semiotic categories: iconic, as in drama, film, and still pictures; indexical, with the sounds effects of films or radio plays; symbolic, as in verbal storytelling; and a combination of iconic and symbolic, as in dialogue, where we have an iconic reproduction of symbolic signs. The iconic and the symbolic modes of signification have narrative power far superior to the indexical, at least when the indexical is used by itself. If we exclude the rare and very lim3 ited case of a story told entirely by noises, narrative, or more precisely narration, comes in two kinds: a mimetic-iconic kind, based on showing, and a diegetic-symbolic kind, based on storytelling. The mimetic kind can imitate verbal exchanges, as in film and drama, or limit itself to the imitation of gestures, as in mime, silent film, still pictures, or series of still pictures; similarly, the diegetic kind may or may not embed an iconic representation of verbal communication, depending on whether or not it uses dialogue. 2. Though imitative, onomatopes are not devoid of conventionality; otherwise, they would be similar across languages. 3. An example of a story told exclusively by indexical noises can be found at https://www. youtube.com/watch?v=7-7eekV9gPc.

114  •  C hapter 6

FIGURE 4. Mimetic and diegetic narratives

Through this model I hope to avoid two weaknesses of Plato’s account. (1) He makes no distinction between three kinds of imitation: dramatic imitation through actors, the imitation of characters through quoted dialogue, and the authorial imitation of first-person narrators. The case of an author imitating the discourse of a first-person narrator can be described as a mimesis of diegesis,4 while the case of fictional dialogue involves an additional level of mediation: an author imitates the diegetic discourse of a narrator; this discourse in turn performs a mimesis of the speech of characters.5 As for dramatic imitation through actors, it differs from quoted dialogue through the absence of a quoting discourse. (2) The claim that the poet speaks for himself in simple diegesis may hold for truth-oriented nonfictional narratives, but in fiction, authors do not assert the propositions they present to the audience, and they are not responsible for their truth (cf. chapter 2). Plato, in other words, was not familiar with the theories of fictionality popular today nor with the widely accepted narratological distinction between author and narrator. My proposal (shown in figure 4) maintains a distinction between the direct imitation of speech inherent in dramatic enactment and the indirect imitation found in the dialogue parts of diegetic narration, where the speech of char 4. The inverse case of a diegesis of mimesis also occurs, for instance when a movie is described by a commentator for blind people. In some countries in Africa, it is common to have movies narrated to the audience during the showing because the movie is in a language the audience cannot understand. 5. This idea that the author imitates the discourse of a narrator applies particularly well to the case of a first-person narrator who openly engages in a communicative act by either speaking or writing, but it is more problematic for the discourse of an impersonal third-person narrator whose communicative status is indeterminate. Needless to say, optional-narrator theories would reject it in this second case.

M imesis and D iegesis  •  115

acters is mediated by a narrator. But just as diegetic narration can embed mimetically reproduced dialogue, mimetic narration can embed diegetic narration, for instance when the characters of a play tell stories to each other. It is therefore necessary to distinguish a primary level from an embedded level in both diegesis and mimesis. From now on, when I speak of mimetic or diegetic narration, I mean the primary level.

Cognitive Implications of Diegesis and Mimesis Through its language-based nature, diegetic narration mediates over distance. It is one of the most powerful semiotic properties of language that it can refer to absent objects; this ability is crucial to verbal storytelling, since it would make no sense to tell audiences about what they can see and hear by themselves. (TV sports broadcasts may appear to be an exception to this claim, but they use diegetic commentary to complement the mimetic images and to stress narrative developments that the spectator could miss.) Insofar as diegesis relies on an act of storytelling, the reader or hearer is aware that the events of the story are not taking place here and now6 but rather are mediated by the discourse of a narrator who recounts what happened in another place and at another time. Moreover, since language does not physically resemble its referents, readers or hearers must use their imagination to form a mental representation of the objects that populate the storyworld. Mimetic narration, by contrast, conveys a sense that characters and events are directly present. The story unfolds here and now, in a simulacrum of real time, and there is no awareness of an act of mediation. I am not saying that events are really unmediated, like events that happen in the real world. Spectators remain conscious in the back of their minds that what is shown is an image made by somebody, but this image is perceived by the senses, and the senses can only perceive what is present, in contrast to the mind, which can imagine absent objects. In mimetic narration, spectators can directly see or hear what exists or happens in the storyworld; the focus of their mental activity is therefore not to imagine characters and events but to interpret their sensory experience as a logically coherent story. While the mimetic mode is often characterized as showing, as opposed to telling, the audience does not imagine that the storyworld is being shown; it simply exists in front of them, just as the world exists around us in real life. This sense of immediacy is expressed by the metaphor of the missing fourth wall to describe the theatrical experience: 6. This can also hold for present-tense narratives: the present is rapidly becoming the preferred narrative tense, and the more widespread it becomes, the more it loses its temporal meaning.

116  •  C hapter 6

spectators are intruders who witness actions and overhear conversations that happen independently of their presence. Communication occurs among the characters and, within the real world, between the spectator and the author, or between the spactator and the actors, but not between a narrator and narratee.

Mimetic Elements in Diegetic Narratives Now that I have defined the basic difference between mimetic and diegetic narration, let me deconstruct it. As I showed in the introduction, languagebased narrative fiction rarely reproduces a genre of factual narrative. When it fails to do so, Phelan’s communicative formula “somebody telling somebody else, on some occasion, and for some purposes, that something happened to someone or something” (Narrative Theory 2012, 3; italics original) is not applicable, because we cannot define the occasion, the somebody, or the purpose. While real-world, factual storytelling presupposes a temporal and spatial distance between the act of narration and the narrated events, this distance can be either emphasized or negated in fiction, depending on whether the text imitates a communicative situation or not. In some eighteenth-century novels (Henry Fielding’s Joseph Andrews or Laurence Sterne’s Tristram Shandy), as well as in some postmodern texts (Samuel Beckett’s The Unnameable), the narrator is so intrusive, and there is so much emphasis of the narrative act, that the telling obscures the told. On the other hand, in some modernist novels, the communicative act disappears from the reader’s consciousness, narrative content becomes the main focus of attention, and, to use a muchquoted formula, the events seem to be telling themselves. Following the lead of Percy Lubbock, critics of the early twentieth century have described these two types of strategies in terms of an opposition between telling and showing (Rabinowitz 2005; Klauk and Köppe, 2014b). As Lubbock explains: “Certainly [Maupassant] is ‘telling’ us things, but they are things so immediate, so perceptible, that the machinery of the telling, by which they reach us, is unno7 ticed; the story appears to be telling itself ” (1921, 62). A number of techniques have been identified as showing, and aspirant fiction writers are advised to favor them over the techniques associated with telling because they provide a more vivid image of the storyworld and are, therefore, more conducive to immersion and illusion than techniques that create distance from the narrated events. First and foremost among these techniques is the use of dialogue, since dialogue constitutes mimetic islands within storytelling; other techniques 7. This idea of the events telling themselves is popular among proponents of the narratoroptional theory, for reasons that are easy to understand.

M imesis and D iegesis  •  117

associated with showing as opposed to telling are scene versus summary; what Franz K. Stanzel ([1979] 1982) calls figural narration (i.e., adopting the point of view of a character) versus objective report; absence versus presence of narratorial intrusions such as evaluative comments and digressions; and free indirect quotation that imitates a character’s speech and thought patterns versus indirect quotation that retains the quoter’s voice. Yet it is difficult to take the telling/showing opposition beyond the level of impressionistic judgment, because there is no such thing as showing through language beside dialogue, if by showing one understands a mimetic representation based on iconicity. As Genette observes, “Showing [in language-based fiction] can only be a way of telling” (1980, 166). The techniques associated with showing are those that background the narrator and the act of storytelling. Typically, a narrative begins as telling, as it posits a narrator who lures the reader with an abstract of the coming tale; then it veers into a showing mode, allowing the reader to become more and more immersed in the story; and it reverts to telling as the narrator concludes the tale with an epilogue or a moral. But some narratives jump right away into a showing mode, for instance the beginning of “Eveline” by James Joyce: “She sat at the window watching the evening invade the avenue. Her head was leaned against the window curtains and in her nostrils was the odour of dusty cretonne. She was tired” ([1914] 2017, 27). While the showing mode tries to simulate through language the directness of mimetic narration, a much more literal appropriation of the mimetic mode by diegetic narration is the phenomenon of multimodal narration. Through the insertion of visual documents such as illustrations, photos, or reproductions of maps, advertisements, newspaper pages, postcards, and handwritten notes, multimodal narrative seeks to overcome the limitations of language concerning the representation of visual information.8 Illustrations facilitate the mental simulation of the story by giving a perceptible appearance to characters and settings, while reproduced documents containing written inscriptions attract attention to the material support of language. These documents are meant to be read, but they differ cognitively from the diegetic text that tells the story in that visual features such the stains on the paper, the typographic arrangement of letters, or, if handwritten, the particularities of the handwriting become significant. Electronic texts are so full of visual data— thanks to the ability of the computer to code images efficiently—that it is often hard to tell whether they are diegetic narratives with lots of mimetic elements or vice versa. 8. Novels that make heavy use of reproduced documents, such as newspaper pages, are Marisha Pessl’s Night Film (2013) and Zachary Dodson’s Bats of the Republic (2015). In S, by J. J. Abrams and Doug Dorst (2013), flat objects are inserted between the pages.

118  •  C hapter 6

Diegetic Elements in Mimetic Media While diegetic narration can increase its imaginative appeal and immersive power through mimetic elements, conversely, the use of diegetic elements can expand the reach of mimetic narration. There are two main reasons for including diegetic elements in a mimetic medium. The first is to overcome the limitations inherent in the medium. Mimetic narrative is often more demanding for the interpreter than the diegetic kind because, as Christina Pawlowitsch argues, in diegetic narrative the narrator can tell you what happens, while in the mimetic kind you must find out for yourself (2020, 337). Insofar as purely mimetic narration is restricted to what can be perceived by the senses of an observer, there are many things it cannot do, or can only do with difficulty: represent thought processes, explain causal relations, spell out motivations for action, accelerate or slow the pace of narration, specify the passing of time between two scenes, identify geographic location, represent events as counterfactual, and so on. This kind of information can be easily conveyed through diegetic elements. For instance, in a movie, verbal messages such as “ten years later” or “New York, 1980” can indicate time and place. Written epilogues telling the audience what happened to characters after the end of the action are a common feature of biopics. Images always present action as happening now, but voice-over narration can situate events in the past. In the movie Amadeus, the voice-over narration is spoken by an aged Salieri confined to a mental institution many years after Mozart’s death. When the screen shows interactions between Mozart and Salieri, the voice-over narration describes events as past, but the image shows them in the present. Another function of voice-over narration is to combine a first-person with a third-person narration. It is not easy for the film medium to present events as experienced by a character,9 because images typically show characters in the third person, and they focus on physical actions rather than on thought processes, but through voice-over narration, we can at the same time see characters from the outside and access their mind from the inside. Another frequent use of diegetic narration in a basically mimetic medium occurs in comics. The image and the speech bubbles that contain dialogue can be considered mimetic, but comics frequently include diegetic storytelling (known as “recitative”). For instance, the opening frames of Maus (figure 5) combine a diegetic narration located above the frame or enclosed in a rectangular box with a mimetic narration that occupies the rest of the frame. In the first frame, the eye moves from the narrative box to the image, and the information in the box provides a basic exposition to the reader of the relationships between the two characters seen

9. On the representation of subjectivity in various media, including film, see Thon (2016).

M imesis and D iegesis  •  119

FIGURE 5. Mimetic and diegetic narration in Maus, by Art Spiegelman

in the image; in the second frame, the reader moves from the mimetic image to the diegetic narration, and the diegetic narration fulfills two roles: first, it presents the private thoughts of the narrator as he encounters his father (“he had aged a lot”), second, it provides the reader background information that would be extremely difficult, if not impossible, to render in the mimetic mode. Another example of the use of diegetic elements to complement mimetic presentation is the use of menus and information about the avatar’s situation in video games. Peter Rabinowitz (2004) has gone as far as suggesting that the orchestral score in the opera, which clearly originates outside the storyworld, provides some kind of diegetic comment on the mimetic action. The second function of diegetic elements in a mimetic medium is to distantiate the spectator from the action by breaking the illusion of an unmediated unfolding of events. This function is far more common in the theater than in film, I assume because the theater cannot rival the expressive resources of film and has therefore been pushed into a self-reflexive mode in order to distinguish itself from its more immersive rival. Also relevant is that film is a popular medium, while the theater, as a highbrow medium, is more open to experimentation. A precedent exists in the chorus of Greek drama, which comments on the enacted plot from a position either internal or external to the storyworld. Monika Fludernik (2008) and Brian Richardson (2001) have provided many examples of an anti-immersive or anti-illusionist use of diegetic elements in the theater of the twentieth century. For instance, the theater may use a narrator figure to present the dramatic action; since this figure appears on stage, its intervention is much more intrusive than the voice-over narration of

120  •  C hapter 6

film, which remains compatible with immersion because it is normally uttered by an invisible narrator and accompanies silent scenes. In the theater, overt narration is not used to comment on an evolving visual scene but, rather, alternates with the speech of characters, so that the diegetic element interrupts the mimetic representation. Another type of diegetic intrusion is the use of written signs instead of mimetic stage settings to indicate the location of a scene, a strategy used in the epic theater of Bertolt Brecht. Stage directions should normally remain invisible in performance, but in what Fludernik calls narrativized stage directions—a feature found in some plays by George Bernard Shaw—they are so detailed that they become impossible to stage mimetically. Should one then read the play silently as written narrative rather than as direction for performance; should the stage directions be ignored or only partially fulfilled; or should they be read aloud during the performance? This last practice is occasionally found in radio drama because it compensates for the lack of visual channel, but it carries a strong anti-immersive, self-reflexive effect.10 Insofar as diegetic elements can penetrate into film or drama, and mimetic elements can be used in language-based narrative, a distinction must be made between mimesis and diegesis as constitutive features of media, and showing and telling as narrative techniques. As Christer Johansson explains: “So the distinction between telling and showing is sometimes a distinction between media, sometimes a distinction between narrative forms and modes, and sometimes a combination of these things, that is, a distinction between narrative forms and how they are realized in different media” (2012, 157; my emphasis). As the words in italics suggest, the possibilities of showing and telling are not really the same in different media. Language-based narrative can show either in a figurative sense, through mediation-effacing techniques, or in a literal sense, through the insertion of visual documents. On the other hand, the mimetic medium of film can tell literally by superposing voice-over narration upon silent images, and drama does so by using a visible narrating agent whose discourse alternates with the mimetically shown action, but I cannot imagine how they could tell figurally.

Conclusion: Experiencing Diegetic and Mimetic Media It is now time to return to the question with which I opened this chapter: what exactly do users imagine in the case of diegetic and mimetic narration? This question cannot be answered with MRI equipment, nor by relying on the lat 10. On theater narratology, see Nünning and Sommer (2008) and Horstmann (2018). On radio drama, see Bernaerts and Mildorf (2021).

M imesis and D iegesis  •  121

est theories of mind: we must depend on introspection, not brain science (at least, not yet). Many fictional narratives begin by putting in place the communicative scheme of a narrator telling a story to an audience. But the more readers get caught up in the story, the more immersed they are in the storyworld, the more the communicative act recedes from their attention, and the more the told eclipses the telling from their minds. This is what early theorists like Percy Lubbock meant when they claimed that in the showing mode the story seems to be telling itself. With some narrative forms—such as third-person omniscient impersonal narration—the communicative situation cannot be naturalized as the imitation of a real-world, truth-centered scheme, and this scheme is therefore irrelevant from the very beginning. I don’t think that the readers of “Eveline,” whose beginning is quoted above, try to imagine a narratorial act of communication. By declaring the scheme irrelevant, I do not mean that the reader tries to apply it and gives up in frustration, but rather that it is not activated at all in the reader’s mind and that its absence does not lead to a conscious judgment that the narrative is “unnatural.” Omniscient narration could never occur in real-world, truth-centered communication because its felicity conditions could never be fulfilled by a human being, and yet the readers of fiction accept it without question, as if it were fully normal. The question of what we imagine in the case of mimetic narration such as theater and film has raised endless debates among philosophers and film theorists (Currie, Bordwell, Matravers), without reaching a consensus, and here I will do no more than make a few observations, without proposing a comprehensive solution. The big difference between language-based diegetic and mimetic media is that while language makes people imagine its referents, mimetic media offer concrete sensory data, and as Derek Matravers (2014) suggests, there is much less to imagine: the spectator’s imaginings concern what cannot be apprehended by the senses, such as private thoughts and motivations, and they do not differ in kind from what people must imagine to interpret real events. The metaphor of the missing fourth wall to describe the experience of the theater presupposes that spectators project themselves as witnesses occupying a specific and fixed location within the space where the action unfolds. This assumption is supported by the fact that the distance between the spectator and the stage is fixed, but the concept of the missing wall cannot be extended to the case of film because, as Gregory Currie observes (1995, 170–78), spectators would have to shift their point of view with each new camera shot, and in the case of close-ups, they would have to situate themselves within a few inches of the characters, a point of view that would never occur in practical life.11 The metaphor of the missing fourth wall 11. This point is also made by Bordwell (1985).

122  •  C hapter 6

may be more appropriate for the theater than for film, but it is ultimately unsatisfactory for both because even though theater and film spectators experience the storyworld as unmediated presence, they do not project themselves into its space as embodied witnesses but rather accept their ability to see and hear what happens in the storyworld without asking questions, just as fiction readers accept without question omniscient narrators and noncommunicative situations. I regard the case of voice-over narration as typical of the cognitive attitude required by the mimetic mode because the dual perspective of voiceover narration, which presents the events as simultaneously past (through the language channel) and present (through the image), demonstrates the futility of a naturalizing explanation. Whereas unnatural narratology would say “look how impossible it is,” and natural narratology might attempt to naturalize voice-over narration, for instance by saying that the image represents the visual memories of the narrator, I believe that the spectator accepts the unnaturalness of voice-over narration naturally, without reflecting on its logical incongruity. This attitude of unconditional acceptance, which Coleridge called a willing suspension of disbelief, is a prerequisite for immersion; but it can be broken by using diegetic elements to frame, question, or even disrupt, rather than to complement the mimetic performance. As the examples from contemporary theater suggest, the intrusion of diegetic elements in a mimetic performance can break the impression of unmediated perception and promote awareness of the true nature of the storyworld, by presenting it as a staged show and not as reality. My examples of combinations of diegetic and mimetic elements expose an asymmetry in narrative practice: while all the cases of insertion of mimetic elements in diegetic narration that I have presented contribute to the immersivity of the text, by allowing a fuller representation of the storyworld, the insertion of diegetic elements in mimetic media can either compensate for the expressive limitations of the mimetic mode, thereby promoting immersion, or, on the contrary, break up aesthetic illusion. Would a broader search turn up illusion-breaking uses of mimetic elements in diegetic narration? I doubt it, because thanks to their sensory nature, mimetic elements have an intrinsic mediation-denying and presence-creating power. But never say never: I will therefore leave the question open.

CHAPTER 7

Parallel Worlds Physics, Narrative, and the Multiverse

The idea that reality—the sum total of what exists—may include worlds other than the world we experience every day ranks near the very top of the topics that fascinate the human mind. We find its manifestations in a wide variety of fields: in medieval cartography, with the representations of fantastic creatures in the terrae incognitae that lie beyond the limits of the known world (Brown [1949] 1977); in philosophy, with Leibniz’s doctrine of the monads; in logic, with the postulation of possible worlds to define the truth conditions of modal operators and counterfactuals (Lewis 1973, Kripke 1963, Rescher 1991); in technology, with the hype that surrounded the development of virtual reality (Ryan 2001); and of course in literature and the visual arts, which did not await the development of science fiction to produce a steady stream of foreign worlds. Last but not least, the idea of multiple realities has made its way into theoretical physics, a development I discuss in the first part of this chapter. In the second part, I turn to narrative manifestations of the idea of multiple worlds, or the multiverse, with special emphasis on narrative inspired by theoretical physics.

Multiple Worlds in Physics In physics the idea of multiple realities is known as the “parallel universes” or as the “many-worlds” interpretation. (World and universe are used inter123

124  •  C hapter 7

changeably.) In my presentation of the notion of parallel universes, I follow a taxonomy proposed by Max Tegmark in Scientific American (2003), but I also rely on work by other authors in my individual discussions of the categories. Tegmark distinguishes four types of parallel universes, which he associates with four levels of the multiverse, the global scheme that encompasses everything that exists. Though Tegmark does not explain the rationale for his ordering of levels, it seems that the higher the level, the more improbable it is that human beings will ever visit its worlds.1

Level 1 The postulation of parallel universes of level 1 rests on three assumptions: 1. Space (or rather, the space that extends along the three classical dimensions)2 is infinite. 2. The observable universe, known as “Hubble space” to astronomers, is limited in size. 3. Matter is not limited to the observable universe. It follows that the outer regions of space could be teeming with galaxies, stars, and planets, in other words, with parallel universes. “According to current theories,” writes Tegmark, “processes early in the big bang spread matter around with a degree of randomness, generating all possible arrangements with non-zero probability” (2003, 42). If space is infinite, it has room for more than all the different possible combinations of elementary particles that make up the observable universe. There is consequently a good chance that the combination that describes our universe is realized more than once, and that we have counterparts of ourselves somewhere in the multiverse. There will also be close copies of our universe in which our counterparts undergo a different fate. According to Tegmark, the multiple universes of level 1 all follow the same laws of physics, but they differ in their initial conditions. They may also travel at different speeds in time and space. Our doppelgängers in the exact copies of our universe may, for instance, already have experienced the year 2030.

1. If it weren’t for the third level, one could also say that the higher the level, the more adjustments are needed in the laws of physics compared with the laws of our world. But level 3 requires no adjustments, as we will see in the section devoted to it. 2. Superstring theory postulates more than three space dimensions (9 or 10), but the additional dimensions are tightly curled up rather than infinite (Greene 1999).

Parallel W orlds  •  125

Level 2 While level 1 parallel universes exist in the same time and space as our universe, those of level 2 are found in alternate spaces. These alternate spaces develop as the result of irregularities in the stretching of space that has continually been taking place since the big bang. According to Tegmark, some regions stop stretching and form bubbles; when these bubbles burst, they give rise to a new level 1 multiverse, complete with its multiple parallel universes. Level 2 universes consequently presuppose a process of recursive embedding within the fabric of space. Brian Greene attributes this recursive embedding to the effect of black holes, those celestial objects whose gravitational field is so intense that it entraps everything that comes close: “Every black hole is the seed for a new universe that erupts into existence like a big bang-like explosion, but is forever hidden from our view by the black hole’s event horizon” (1999, 369). Greene and Tegmark agree that the universes that come into existence as a result of such violent events within the fabric of space may differ from those of level 1 in much more radical ways than level 1 universes differ between themselves. Greene suggests that they may have a different number of elementary particles, or that their space may have a different number of dimensions, while Tegmark believes that they may differ in the value of constants but follow the “same” laws of physics: you just plug different numbers into the same equations.

Level 4 But what if parallel universes followed entirely different laws? Tegmark makes room for this possibility in his level 4 universes, which I describe here before those of type 3 because I want to end up with the type that plays the most revolutionary role in theoretical physics. Universes of level 4 are the products of a thought experiment based on the belief that there is no absolute necessity for the laws of physics to be the way they are. “How about a universe,” asks Tegmark, “that obeys the laws of classical physics, with no quantum effects? How about time that comes in discrete steps, as for computers, instead of being continuous? How about a universe that is simply an empty dodecahedron?” (2003, 49). Adopting a Platonic view, by which mathematical structures embody “the ultimate reality,” Tegmark suggests that level 4 universes are the realizations of all the structures that can be mathematically described. The only laws that level 4 universes cannot breach are those of mathemati-

126  •  C hapter 7

cal logic, such as excluded middle and noncontradiction. “If the universe is inherently mathematical,” asks Tegmark, “then why was only one of the many mathematical structures singled out to describe our universe?” (50).

Level 3 The last type of parallel universe postulated by physicists is an attempt to explain the notoriously strange phenomena that take place on the subatomic level—the phenomena of quantum mechanics. In contrast to the worlds of levels 2 and 4, the so-called many-worlds interpretation of quantum phenomena does not presuppose any modification of the laws of physics. “It is ironic,” writes Tegmark, “that level III is the one that has drawn the most fire in the past decade, because it is the only one that adds no qualitatively new types of universes” (2003, 51). The many-worlds interpretation of quantum physics is an attempt to deal with the well-established fact that the behavior of subatomic particles, such as electrons, cannot be predicted with absolute certainty. As Brian Greene writes: “If a particular experiment involving an electron is repeated over and over again in an absolutely identical manner, the same answer for, say, the measured position of an electron will not be found over and over again” (1999, 107). We can determine the probability of an electron (or any other kind of subatomic particle) being found in a specific location, but we cannot predict which one of these possibilities will be actualized. Such probabilities are predicted by using a mathematical object known as the wave function of the particle, and it is captured, for the electron, by the famous Schrödinger equation. Yet of all the possible positions or trajectories of the electron, only one will be realized. Or at least, this is the orthodox interpretation, generally associated with Niels Bohr and the Copenhagen school of quantum mechanics. According to the so-called Copenhagen interpretation, the wave function collapses when an observation is made, so that each elementary particle will be found in a specific location rather than in all the possible states predicted by the wave function. But before the observation takes place, nothing meaningful can be said about the electron, either because it exists in a superposition of all possible states, or because it has neither a velocity nor a position. Schrödinger, who thought the idea absurd (Kaku 2005, 138; Bruce 2004, 59), exposed the paradox through a celebrated thought experiment. A cat is sealed in a box together with an atom of uranium that has a 50/50 chance of decaying in a certain time span. Being a quantum phenomenon, the decay cannot be predicted. Now imagine that if the decay takes place, a mechanism will

Parallel W orlds  •  127

be triggered that kills the cat. Before you open the box, the atom will have both decayed and not decayed, and, in the interpretation that Schrödinger wanted to discredit, the cat will be dead and alive at the same time: “In order to describe the cat, physicists add the wave function of the live cat and the dead cat—that is, we put the cat in a nether world of being 50 percent dead and 50 percent alive simultaneously” (Kaku 2005, 158). It is the action of opening the door that sends the cat either to Hades or back to her favorite pillow. “But why,” asks Brian Greene, “should fundamental physics be so closely tied to human awareness? If we were not here to observe the world, would wave functions never collapse? . . . Was the universe a vastly different place before human consciousness evolved on planet earth?” (2004, 207). For the proponents of multiple universes, the logical contradiction, as well as the problematic notion of the decisive role of the observer, is unacceptable. As Tegmark writes: Many legitimate wave functions correspond to counterintuitive situations, such as a cat being dead and alive at the same time, in a so-called superposition. In the 1920s physicists explained away the weirdness by postulating that the wave function “collapsed” into some definite classical outcome whenever someone made an observation. This add-on had the virtue of explaining observations, but it turned an elegant theory into a kludgy, non-unitary one .  .  . Over the years many physicists have abandoned this view in favor of one developed in 1957 by Princeton graduate student Hugh Everett III. He showed that the collapse postulate is unnecessary. Unadulterated quantum theory does not, in fact, pose any contradiction. (2003, 46)

What Everett suggested is that random quantum processes cause the universe to split into multiple copies, one for each possible outcome. The cat, consequently, is dead in one parallel universe, alive in another, and watched in each universe by different copies of the observer or by none at all. In this interpretation, the multiple possibilities described by the wave functions are more than possibilities—they are all actual phenomena taking place in different worlds. The parallel-worlds interpretation of quantum physics implies that every quantum phenomenon, this is to say, every movement of subatomic particles, causes the universe to split—the particles taking a different route in each of the resulting parallel universes. Wittily captured by the title of Colin Bruce’s book Schrödinger’s Rabbits: The Many-Worlds of Quantum, this absolutely dizzy­ing explosion explains why the idea of parallel universes has encountered far stronger opposition for level 3 than for levels 1 and 2, where the birth

128  •  C hapter 7

of other worlds remains a special event rather than a routine phenomenon that takes place every split second. (Level 4 is far too speculative to provoke passionate reactions.) The idea of a constantly splitting universe violates the principle of Occam’s razor, which prohibits the introduction of unnecessary entities into a theory, but once we assume the existence of parallel universes, postulating an infinity of them will not further violate Occam’s razor, because what the principle arguably seeks to limit is a qualitative, rather than quantitative proliferation of theoretical entities. The entity that violates Occam’s razor is the first additional world; all the others come for free. If physicists agree on one issue, it is that elementary particles seem to obey different laws than large-scale phenomena. The many-worlds interpretation of quantum phenomena is one of the many attempts to conquer what Brian Greene calls the holy grail of theoretical physics: devising a theory that reconciles quantum mechanics with Einstein’s relativity, two theories which “cannot be both right” (1999, 3; italics original) in their current formulation.

Multiple Worlds in Philosophy and Narratology The philosophical counterpart of the parallel worlds of cosmology and theoretical physics is possible worlds theory. I refer the reader to the presentation of the theory in the introduction. As we have seen, a fundamental tenet of PW theory is the distinction between one actual world and many merely possible ones. According to David Lewis’s doctrine of modal realism, all possible worlds are equally real, and all possibilities are realized in some world, independently of whether somebody in our world thinks of them or not. Lewis reconciles the idea that all possible worlds are real with the opposition of one actual world to many possible ones by distinguishing the concepts of reality and actuality and by conceiving actuality as a matter of point of view: “The actual world” means “the world where I am located,” and all possible worlds are actual from the point of view of their inhabitants (1979, 184). The narratological applications of PW theory support this conception of actuality, because it explains how fictional universes can be centered around their own actual world, even though, from the point of view of our world, they are nonactual creations of the imagination. To the reader immersed in a fiction, the center of the fictional universe functions temporarily as the actual world, and this reader relates to the characters as if they were flesh-and-blood human beings. Physics, by contrast, has little use for the distinction between the actual and the nonactual, and consequently for any definition of actuality. If parallel universes exist, they are collections of material objects, such as stars, planets, and galaxies, and they are not structured by the modal opposi-

Parallel W orlds  •  129

tion between a single actual components and many nonactual possible ones. Within parallel universes, all objects exist in the same ontological mode—the mode of reality. Even when physics deals with probabilities—that is, when it calculates the wave function of a particle—the many-worlds interpretation denies any contrast between the actual and the nonactual, since it regards all probabilities as simultaneously realized in some world. “In the Many Worlds Approach,” writes Greene, “. . . no potential outcome remains merely a potential . . . Every potential outcome comes out in one of the parallel universes” (2004, 206). While physics is indifferent to Lewis’s conception of actuality, both Tegmark and David Deutsch, another strong proponent of the many-worlds interpretation of quantum phenomena, have sought interdisciplinary support in Lewis’s postulation of the objective existence of multiple worlds. According to Deutsch, The fruitfulness of the multiverse theory in contributing to the solution of long-standing philosophical problems is so great that it would be worth adopting even if there were no physical evidence for it at all. Indeed, the philosopher David Lewis, in his book On the Plurality of Worlds, has postulated the existence of a multiverse for philosophical reasons alone. (1997, 340)

As for Tegmark (2003, 50), he invokes Lewis in support of the idea that underlies the fourth level of parallel universes: all mathematical structures exist physically. Lewis (2004), conversely, justifies his belief in multiple realities by arguing that the notion of quantum collapse in physics is untenable. For physicists, the existence of alternate worlds must be subject to falsification or verification. Tegmark (1993, 51) takes great pains trying to convince his reader that the existence of parallel universes is not mere fantasizing but an experimentally testable theory, and Bruce (2004, 140–74) mentions several experiments that supposedly demonstrate the existence of many-worlds. The narratological application of PW theory, by contrast, is largely indifferent to the question of modal realism. For narratologists, possible worlds can be treated as what is known in philosophy as a theoretical fiction (an imaginary entity postulated for its explanatory power), and as long as their applications are found to be useful, it does not matter whether they exist objectively or not.

Narrating the Multiverse When we read a narrative, we are naturally inclined to regard its universe as centered around one, and one only, actual world, and it takes a deliberate

130  •  C hapter 7

effort on the part of the author to fight this tendency of the reader. But how can a narrative suggest the idea of parallel universes existing objectively? Jorge Luis Borges has provided a description of a multiverse narrative where all possibilities are realized in his short story “The Garden of Forking Paths.” This story describes a fictional novel, also called “The Garden of Forking Paths,” written by the Chinese sage Ts’ui Pen, a character in the main story.3 To the narrator (a descendant of Ts’ui Pen), “the book is a contradictory jumble of irresolute drafts. I once examined it myself: in the third chapter the hero dies, yet in the fourth he is alive again” (Borges 1999, 124). The scholar Stephen Albert explains to the narrator the reason for the apparent contradictions: In all fictions, each time a man meets alternatives, he chooses one and eliminates the others; in the work of the virtually impossible-to-disentangle Ts’ui Pen, the character chooses—simultaneously—all of them. He creates, in this way, “several futures,” several times, which themselves proliferate and fork. This is the explanation for the novel’s contradictions. Fang, let us say, has a secret; a stranger knocks at his door; Fang decides to kill him. Naturally, there are various possible outcomes—Fang can kill the intruder, the intruder can kill Fang, they can both live, they can both be killed, and so on. In Ts’ui Pen’s novel, all the outcomes in fact occur; each is the starting point for further bifurcations. Once in a while, the paths of the labyrinth converge: for example, you come to this house, but in one of the possible pasts you are my enemy, in another, my friend. (125; italics original)

Relying on a metaphorical process fundamental to human cognition—representing time in spatial terms—Borges invokes the image of the forking paths to describe a fundamentally temporal phenomenon: Unlike Newton and Schopenhauer, [Ts’ui Pen] did not believe in a uniform and absolute time; he believed in an infinite series of times, a growing, dizzy­ ing web of divergent, convergent and parallel times. That fabric of times that approach one another, fork, are snipped off, or are simply unknown for centuries, contains all possibilities. In most of those times, we do not exist; in some you exist but I do not; in others, I do and you do not; in still others, we both do. In this one, which the favored hand of chance has dealt me, you have come to my home; in another, when you come through my garden you 3. The structure of the embedded “Garden of Forking Paths” does not describe the “Garden of Forking Paths” of the main level, since the main-level story is standard narrative with no forks.

Parallel W orlds  •  131

find me dead; in another I say these same words, but I am an error, a ghost. (127; italics original)

The reason why Ts’ui Pen’s novel looks like a jumble of contradictions lies in the merging of paths. It is only in space that the paths of two visitors arriving at the same house can come together. But in Borges’s text the forking paths are a metaphor for time, and timelines cannot merge, because this would mean that the propositions “x is y’s friend” and “x is y’s enemy” would both be true in the biography of x, in violation of the law of noncontradiction.4 The merging of distinct paths is also problematic from the point of view of narrative motivation, because x and y’s personal relationship is determined in the past, and the past casts a shadow on the future. The world where x and y are friends gives access to a different set of states, and consequently opens different narrative possibilities, than the world in which they are enemies. PW theory would deal with this situation by saying that x does not arrive at the same house in the same world through different but converging timelines but, rather, that x reaches different houses in different worlds that occupy corresponding spatial coordinates within their respective world. For a Borgesian “Garden of Forking Paths” to maintain logical coherence along each path, it should consequently be modeled on a decision tree whose branches never merge, rather than on a network diagram allowing circles and loops. But if the branches are kept neatly separate, the worlds they represent will decohere, to borrow a physics term, and Tsui Pen’s novel will be a collection of separate stories rather than a unified narrative multiverse. Just as it takes an entanglement of destiny lines to turn the lives of distinct individuals into a plot, it takes interference between its constituent worlds to turn a fiction with a multiverse cosmology into a cohesive narrative. This may be why Borges envisions a merging of paths, despite the logical contradiction inherent in the phenomenon. Alan Lightman’s Einstein’s Dreams offers an example of a multiverse fiction without world entanglement. One of the chapters represents three worlds that branch from a common stem. Each of these worlds contains counterparts 4. Philosophers concerned with time logic, such as A. N. Prior, consider the possibility of a splitting of timelines, but not of their merging, because “the future has openness to alternatives which the past has not” (Prior 2003, 57). According to David Lewis, some philosophers even object to the idea of splitting, “for when one becomes two, it seems that one single thing is identical to two different things” (2004, 12). Lewis’s solution to this problem is that “there are two all along, though before the branching the two were temporarily identical in the sense that they shared an initial temporal segment” (12). Here “sharing a temporal segment” means being part of different timelines that follow parallel courses, rather than being part of the same timeline. As Lewis explains elsewhere, “There is an initial segment of one world, and there is an initial segment of the other, which are perfect duplicates” (1986a, 70).

132  •  C hapter 7

of the same individuals: a man living in Berne and a woman he once met in Fribourg. In the first world, the man “decides not to see her again. She is manipulative and judgmental, and she would make his life miserable. Then, in three years, he meets another woman in a clothing shop in Neuchâtel . . . After a year she comes to live with him in Berne” (1993, 19–20). “In the second world, the man . . . decides he must see the Fribourg woman again . . . They make love, loudly and with passion. She persuades him to move to Fribourg. He leaves his job in Berne and begins to work at the Fribourg Post Bureau” (20). “In the third world, he also decides that he must see her again . . . After an hour, she says that she must leave to help a friend, she says goodbye to him, they shake hands. He travels the thirty kilometers back to Berne” (21). We are told that all three chains happen simultaneously in the multiple dimensions of time, but there is no communication from branch to branch, and what happens in one does not affect the others. Because of this lack of interference, the three chains create different stories rather than a unified plot encompassing parallel realities. But Lightman’s book is designed as a series of autonomous “dreams” illustrating the diverse conceptions of time that the imagination can extrapolate from Einstein’s theory of relativity, and this fragmentation, which takes place within one of the dreams, is fully compatible with the thematic unity of the book as a whole. For a text to impose a multiverse cosmology, it must be based on a decision tree or on a diagram with parallel branches (for those texts that do not assign a common origin and a genealogy to their component worlds), and all the branches must possess equal ontological status. But this is not sufficient to create situations of narrative entanglement. In order to do so, the text must not only move up and down along the branches, it must also perform lateral jumps from branch to branch, and there should be a consciousness within the narrative multiverse that is aware of the jumping. In Lightman’s text, narration follows one branch to its end; then backtracks in time, takes another branch, follows it and backtracks once again to take the third branch, but the characters along the three branches do not know that they have counterparts, and the various stories do not influence each other. In a fully integrated multiverse narrative, by contrast, characters either travel physically from branch to branch or know with certainty that other branches exist objectively. This knowledge affects their behavior and, consequently, alters the history of their own universe.

Classical Forms of Plural Ontology By classical forms of plural ontology, I mean narratives that raise the issue of multiple realities, but without endorsing a genuine multiverse cosmology

Parallel W orlds  •  133

inspired by the many-worlds interpretation of quantum physics. These forms are the narrative of transworld exploration, the narrative of alternate history, and the time-travel narrative. In most narratives of transworld exploration, such as Gulliver’s Travels, Star Trek, and Star Wars, all the worlds belong to the same spacetime continuum, and the travels of characters are presented as nothing more than a journey through an extended space, even when this journey requires vehicles far beyond current technology. In Gulliver’s Travels, the continuity of spacetime is demonstrated by maps that locate the various worlds within real-world geography: for instance, Lilliput is near Sumatra, and Brobdingnag is a peninsula on the California coast. But in a variant that I call the wormhole narrative, transworld travel is limited to narrow portals usually open only to those with special powers (supernatural abilities, birthright, being selected among all individuals by the masters of the other world, etc.), and passing through these portals amounts to being transported into an alternative reality. These narratives usually present what Pavel (1986a, 57) calls a salient structure, or dual ontology: two realms differing sharply as to what is possible in them, such as the sacred and the profane. In the first book of C. S. Lewis’s Chronicles of Narnia, for instance, children pass through a wardrobe in the house of an old professor and end up in the magical realm of Narnia, where they meet supernatural creatures like Aslan the Lion and his antagonist the White Witch. In the fairy tale Jack and the Beanstalk, Jack climbs what looks like an ordinary plant and discovers that the leaves support entire worlds. And in Harry Potter, the wall at platform 9 3/4 of the King’s Cross railway station in London serves as a portal between the world of everyday life, where Harry is a poor orphan mistreated by his adoptive family, and the magic world of the Hogwarts School of Witchcraft and Wizardry, where he learns to be a wizard. Whether this type of narrative rests on a two-world ontology with backand-forth movement, or on a network of interconnected worlds (Dannenberg 1998), it presupposes a discontinuous spacetime pierced with secret passageways, similar to the wormholes in the fabric of the cosmos that connect the universes of level 2 (cf. Kaku 2005, 401, and Greene 2004, 264–65). Alternate (or counterfactual) history fiction creates a world whose evolution, following a certain event, diverges from what we regard as actual history. The cause of the divergence may be either a key decision at a special moment by an important historical figure (Napoleon, Hitler, and Roosevelt are favorites) or a small random event starting a causal chain that leads to enormous consequences, in what is known in chaos theory as the butterfly effect. Whether alternate-history fiction presents the fate of the world as determined by human decisions at certain strategic points, or shows it to be the product of forces too numerous and too complex to be controlled, the pur-

134  •  C hapter 7

pose of such thought experiments is to invite reflection on the mechanisms of history, and the real world always serves as an implicit background. But this does not mean that alternate history necessarily proposes a multiple-worlds cosmology, as Philip Roth’s The Plot Against America demonstrates. The Plot Against America explores what would have happened if in 1940 Charles Lindbergh, a Nazi sympathizer, had been elected US president instead of Franklin Roosevelt, but there is no suggestion within the text of the existence of another objectively existing reality. This is not to say that counterfactual history is incompatible with the multiverse cosmology: as we will see in the next section, John Wyndham’s “Random Quest” presents two versions of history as objectively implemented in different worlds. Drawing multiple lines of alternate history is indeed a favorite device of multiverse narrative. Time-travel narratives present a two-way movement along one of the branches of the tree of historical possibilities. But this movement does not imply the existence of multiple parallel universes. In H. G. Wells’s The Time Machine, for instance, the hero travels from Victorian England to a distant future, and then returns to the unchanged world of his spatiotemporal point of departure. Wells’s chrononaut remains a relatively uninvolved observer of social structures, and when he returns to the present he does not use the knowledge acquired during his trip to prevent the gruesome exploitation of the Eloi by the Morlocks observed in the future. But the time-travel story is a genre notoriously riddled with paradoxes (Nahin 1999), and the most important of these paradoxes is rooted in the single-universe cosmology. In the socalled grandfather paradox, you enter a time machine, travel back in time, and kill your grandfather. This means that you will never be born and will never enter the time machine. Then your grandfather will not be killed and you will be born after all, and so on ad infinitum. As Lewis (1986b) and Deutsch and Lockwood (1994) have argued, this vicious circle can only be avoided if one assumes that the actions of the chrononaut create a splitting of worlds, by which the grandfather is killed in one branch and lives in the other. By causing the death of his grandfather, the time traveler puts himself on a timeline in which he will never exist, and he will have to perform a lateral jump to another branch, as well as a jump in time, to return to his native reality. If this sounds almost as paradoxical as the vicious circle, think of the chrononaut as some kind of alien visiting a foreign world for a limited time. The combination of time travel and many-worlds cosmology is illustrated by the film The Butterfly Effect (2004). Evan, the hero, is a young man who experienced some traumatic events as a child: witnessing the suicide of his father, a patient in a mental institution; being forced to act in a pornographic movie by a neighbor; watching a friend torture and kill a dog; and participat-

Parallel W orlds  •  135

ing in a deadly prank with other children. During each of these events Evan blacks out and represses the memories. But many years later, as a college student, he discovers childhood diaries that a therapist made him keep to avoid a total loss of memory, and, upon reading them, he is transported back to the scene of the traumatic events. This gives him a chance to act differently and to change his life, but each time he travels back to the past and chooses a different course of action, unforeseen side effects lead to tragedies that affect other people, until he reaches a world where he dies at birth—sadly the best of all possible worlds.5 What makes The Butterfly Effect a genuine many-worlds narrative is the fact that, as he first returns to the past, and then is carried forward by the flow of time, Evan never returns to the same present from which he came. His time travel takes ever new branches.

Multiverse Narratives and Counterpart Relations From a narratological and philosophical point of view, the most intriguing aspect of the many-worlds model is the idea that we exist in multiple copies— that we are linked by what Lewis (1986a) calls counterpart relations to “other” individuals inhabiting parallel worlds. In this section I propose to explore several narrative variations on the notion of counterpart-hood, focusing on the personal relations of characters to those individuals who are at the same time themselves and somebody else, as well as on the plot possibilities that these variations create. In one multiverse scenario, individuals are represented in each world by a material body, and cross-world travelers enter their counterpart’s body as if it were a hollow envelope, while the regular inhabitant of the body conveniently disappears for the time of the visit (figure 6, left). This creates a mind/body split, since the body of the cross-world traveler belongs to the new world, while his knowledge, memory, and sense of identity are those of his old-world 6 counterpart. In John Wyndham’s “Random Quest,” a young physicist, Colin Trafford, is transported, after an explosion caused by an experiment in nuclear fission, into a version of 1954 England where World War II did not happen, Winston Churchill is not prime minister, and theoretical physics is far less advanced, since it is not supported by the war effort, as it is in the world Trafford comes from as well as in the actual world. In this world Trafford has 5. I am describing here the director’s version available on the DVD. I am told that the version shown in movie theaters had a different ending. 6. I am indebted to Hilary Dannenberg for mentioning to me the two Wyndham stories discussed here.

136  •  C hapter 7

become a very successful novelist. The “original” Trafford (i.e., the one from whose point of view the story is presented, henceforth Trafford I) finds himself in a body partly familiar, partly foreign: while he recognizes his face in a mirror, this face wears a moustache, and his clothes are different from those he wore before the explosion. The involuntary character of the cross-world journey and the ensuing epistemological conflict create narrative situations particularly well suited to arousing the reader’s curiosity. In episodes of powerful tellability, the story describes Trafford’s puzzlement at the changes in his body; the gradual realization that this partly foreign, partly familiar body belongs to another world; the exploration of the new reality; and, through a series of encounters with people who take him as Trafford II, the discovery of the personality of his counterpart, whom Trafford I finds rather repulsive: as a writer, Trafford II caters to popular taste rather than maintaining artistic integrity, and as a husband, he cheats on his wife, Ottilie, née Harshom. When Trafford I visits the flat of Trafford II and meets Ottilie, he realizes that she is the woman he was meant to love and marry, and he spends blissful weeks with her, while his body in world I lies unconscious in the hospital. (We don’t know what happens to Trafford II while Trafford I occupies his body.) Trafford I finally awakens from his coma, and engages in a desperate search for Ottilie’s counterpart in world I. He visits a certain Dr. Harshom to find out whether he knows of any Ottilie Harshom, but the doctor discourages him: the only Ottilie Harshom he knows of belongs to a previous generation; and even if Ottilie existed in world I, she would be a completely different person: Your counterpart varied from you, you say. Well, her counterpart if she existed would have had an entirely different upbringing in different circumstances from the other; the probability is that there would only be the most superficial resemblance .  .  . Somewhere in the back of your mind you are giving house-room to the proposition that unlike causes can produce like results. Throw it out. (1965, 165)

Against Dr. Harshom’s rigid determinism, Trafford clings to a theory that leaves room for randomness. He speculates that the existence of multiple worlds is due to a natural splitting of atoms that takes place all the time and generates all possibilities. (By contrast, the cross-world travel that Trafford experiences is due to a partly human cause, namely a scientific experiment in atomic fission.) I call the type of reality model on which “Random Quest” is based a quantum cosmology because it involves an infinite number of worlds born at every moment, as opposed to those multiverse cosmologies in which

Parallel W orlds  •  137

the sprouting of worlds is due to isolated events, as in Tegmark’s level 2. Here is Wyndham’s description of the quantum cosmology: Perhaps chance is continually causing two different outcomes so that in a dimension we cannot perceive there is an infinite number of planes, some so close to our own and so recently split off that they vary only in minor details, others vastly different. Planes on which some misadventure caused Alexander to be beaten by the Persians, Scipio to fall before Hannibal, Caesar to stay beyond the Rubicon; infinite, infinite planes of the random split and resplit by the random. Who can tell? But, now that we know the Universe for a random place, why not? (159–60)

If all possibilities are realized in some world, there is an infinitesimal chance that the Ottilie of world I has retained enough of her world II personality to be the love of Trafford’s life. Through a fairy-tale ending, this is indeed what happens: Trafford finds the counterpart of the Ottilie of world II in a woman named Belinda Gale who lives in world I. Differences in the circumstances of her birth and upbringing have resulted in a different name and slightly different personalities, but Ottilie/Belinda has maintained a far more constant character than the two versions of Trafford: “Just one distant sight of her was enough. I couldn’t have mistaken [Belinda] for Ottilie but she was so very, very nearly Ottilie that I would have known her among ten thousands” (169; italics original). Rather than entering the corporeal shell of his counterpart, the visitor from a parallel world could confront a fully embodied version of his alter ego (figure 6, center). One of the few authors who have explored the narrative potential of this situation is once again John Wyndham. In his short story “Opposite Number,” Wyndham invents the notion of an “atom of time” to explain the incessant splitting of worlds and individuals: So, here we have Peter Ruddle. An instant later, that atom of time in which he exists is split, and so there are two Peter Ruddles, slightly diverging. Both these time-atoms split, and there are four Peter Ruddles. A third instant, and there are eight, then sixteen, then thirty-two. Very shortly there are thousands of Peter Ruddles. And because the diversion must actually occur many, many times in a second, there is an infinite number of Peter Ruddles, all originally similar, but all different by forces of circumstances, and all inhabiting different worlds—imperceptibly different, or widely different; that depends chiefly on the distance between the original point of fission. (1959, 130)

138  •  C hapter 7

FIGURE 6. Three types of counterpart relations in many-worlds narratives

“Opposite Number” confronts two fairly close versions of Peter Ruddle. Before their splitting apart, Peter Ruddle is a physicist working in an institute with old Whetstone, a mad-scientist character who stubbornly believes in the quantum cosmology and in the possibility of cross-world travel. Peter is in love with Jean, the daughter of old Whetstone. In one world, he marries Jean and pursues Whetstone’s research after the death of the old man and eventually invents a cross-world travel machine, which he uses to visit his counterpart. In the other world, Peter abandons Whetstone for mainstream research projects, and after Jean marries one Freddy Tallboy, he resigns himself to marrying another woman. I’ll call this version Peter I, because the story is told from his point of view. Peter I provides a more interesting narrative perspective than Peter II because he knows less than his counterpart, and the story can focus on his surprise when he is visited in world I by Peter II. The latter’s explanations serve the double purpose of satisfying the reader’s curiosity and of solving Peter I’s puzzlement, as he sees his old flame, Jean, visiting old Whetstone’s lab with a man who looks strangely like himself. The motivation for Peter II’s visit to Peter I is to improve the cross-world travel machine, which works so far only between close worlds, by pooling their intelligences: “You know how one’s mind tends to work in a groove— well, it occurred to me that if I could start one of my ‘doubles’ working on this thing, too, it might lead to a better understanding of it” (131). As the two Peters compare their destinies, it becomes evident that Peter II has chosen

Parallel W orlds  •  139

the right path, and has been rewarded by marrying his true love, while Peter I has betrayed both science and the woman who was meant for him. The sense that counterparts of Peter and Jean were predestined to love each other in all the worlds in which they exist is brought to a moving climax when the Jean of world II feels sorry for her counterpart and urges Peter I to renew relations with Jean I. But this melancholic meditation on missed opportunities morphs into a comedy of errors when Peter II and Jean II are seen holding hands in public and the rumor spreads across world I that Peter Ruddle is cheating on his wife. (Remember that Peter I has married somebody else.) The existential theme of the confrontation of selves has given way to the classic plot device of mistaken identity. Yet another possible type of relationship between counterparts is the merging of their consciousness within a single body (figure 6, right). Here individuals exist in only one copy in each world, but the memory of the traveling character reaches back into both worlds and creates a split identity. An example of this situation is found in Ursula K. Le Guin’s fascinating science fiction novel The Lathe of Heaven. The action of the novel, which was written in 1970, takes place in Portland, Oregon, around 1998. It tells the story of a young man named George Orr who is affected by a strange mental condition: he dreams “effective dreams” that actually change the circumstances of his life. When George wakes up, he finds himself in a new reality that conforms to the dream. He visits a psychoanalyst, Dr. Haber, in the hope of being cured of his condition. The doctor’s treatment consists of hooking George up to a machine that places him in a sleep state, and of controlling his dreams through hypnosis. Haber soon discovers that he can change the world, or, if one interprets the novel according to the plural cosmology, create a new world, by scripting his patient’s dreams. But each time he tries to solve one of the world’s problems by suggesting a dream topic, the dream takes on a nightmarish quality, and the improvements it creates are the side effects of more catastrophic events: for instance, when Haber suggests that George dream a solution to the problem of overcrowding, the dream lands them in a world whose population has been decimated by a pandemic; when he suggests that George dream the end of war between humans, the countries of the post-dream world no longer fight each other because they are all threatened by an invasion of aliens who have already landed on the moon; and when Dr. Haber induces in George a dream in which the aliens have left the moon, they conquer the earth in the post-dream world. (Fortunately, they turn out to be quite willing to cohabit peacefully with humans.) When Haber thinks that he can now self-induce effective dreams, he cures George by having him dream the end of his condition, and he hooks himself up to the sleep-inducing machine; but George pulls

140  •  C hapter 7

the plug, and destroys Haber’s ability to control the world. In the next world, Haber is a patient in a mental institution. The Lathe of Heaven does not explicitly state that George and Haber are transported after each dream into a parallel world, and it is left to the reader either to construct a cosmology that supports the phenomena described in the novel or to leave these phenomena unexplained. When George says that reality is simply being “changed out from under us, replaced, renewed” (1971, 71), he suggests acceptance of the irrational, since this would involve a rewriting of the past, a phenomenon that most philosophers consider paradoxical. (For more on this topic, see my discussion of Emmanuel Carrère’s La Moustache in chapter 8.) Haber, on the other hand, calls the new realities “another continuum” (115), an interpretation strongly suggestive of a transfer to a parallel world with a different past. In a many-worlds cosmology, the dreams fuse George with the George of an already existing timeline, rather than performing a logic-defying rewriting of his personal history. The many-worlds cosmology explains why, after each transition, George’s memory reflects both the world after the dream and the world before the dream: “Although his memory assured him he had held this position for five years now [in the new world], he disbelieved his memory; the job had no reality to him. It was not work he had to do. It was not his job” (146). It is this split memory that motivates George to try to put an end to his effective dreaming; for if he only remembered the new world, he would not know that he came from another reality, and he would not be aware of his responsibility for the present state of affairs. Each dream, each transfer to a new world, adds a new thread to George’s memory, until he becomes unable to keep these threads apart: “He had no idea what, or how much [Heather, his girlfriend] recalled, nor how to fit it within his multiple memories” (173). In the last world, George is finally freed from the terrible responsibility of effective dreaming, but the accumulation of memories from previous lives leaves almost no room in his mind for the formation of a new strand. With his consciousness turned into a battlefield of conflicting memories that tie him to other worlds, George has become totally alienated from his present environment: “The continuity which had always held between the worlds or timelines of Orr’s dreaming had now been broken. Chaos had entered in. He had few and incoherent memories of this existence he was now in; almost all he knew came from other memories, other dreamtimes” (167–68). But the novel ends on a more positive note, as George takes Heather for a coffee break in order to sort out and put back together their scattered memories.

Parallel W orlds  •  141

Philosophical Implications of the Quantum Cosmology A very short story by Larry Niven, “All the Myriad Ways” (1971), proposes a lively dramatization of the philosophical questions raised by the idea that we possess an infinity of counterparts in alternate worlds, and that through these counterparts, we take all the courses of action available to us. The story opens with a textbook exposition of the quantum cosmology: There were timelines branching and branching, a megauniverse of universes, millions more every minute. Billions? Trillions? Trimble didn’t understand the theory, though God knows he’d tried. The universe split every time someone made a decision. Split, so that every decision ever made could go both ways. Every choice made by every man, woman and child on Earth was reversed in the universe next door. It was enough to confuse any citizen, let alone Detective-Lieutenant Gene Trimble, who had other problems. (1971, 1)

The problems that intrigue Trimble are senseless suicide and senseless crime, and they have both developed since the creation of a corporation named Crosstime that specializes in sending missions to parallel worlds and bringing back profitable inventions. Trimble observes a particularly high suicide rate among the pilots of Crosstime, as well as a wave of gratuitous murders in the population at large. The action in the story is provided by Trimble’s investigation of the suicide of Ambrose Harmon, the founder of Crosstime, who jumped from the window of a twenty-ninth-floor apartment after winning a large sum playing poker. But this investigation does not lead anywhere. Trimble discovers the explanation for the suicide not by inspecting the death scene but by reflecting on the logical, moral, and existential consequences of the quantum cosmology. The suicide candidates know that if all possibilities must be realized, the suicide attempt will fail in some possible world, and at least one of their counterparts will survive. Many others will die, but since death is the absolute end of consciousness, their tragedy will not matter to the survivor, who will be entirely focused on the life (or lives) that lie ahead of him. The story ends with Trimble testing the hypothesis on himself, with the expected multiple consequences: And he picked the gun off the newspapers, put it to his head and fired. The hammer fell on an empty chamber. fired. The gun jerked up and blasted a hole in the ceiling fired. The bullet tore a furrow in his scalp took off the top of his head. (11)

142  •  C hapter 7

In a posthumously published paper whose topic eerily dovetails with that of Niven’s story (though there is no evidence that he was familiar with it), David Lewis (2004) argues that in a quantum cosmology, death is impossible. Even if it appears in this world to be a biological necessity, we cannot rule out the possibility that a technological invention will prolong our life forever in another world. In the many-worlds interpretation of quantum mechanics, there will always be a world in which Schrödinger’s cat survives the experiment. But to Lewis, this is not cause for rejoicing, because among “all the myriad ways” of surviving, some will involve a bullet tearing a furrow in our scalp—as Trimble will soon discover—and leaving us in a permanent vegetative state. In an almost lyrical passage made deeply poignant to the reader by the thought that it could represent his last published words (he died unexpectedly shortly after presenting the paper at a conference), Lewis writes: Everett’s idea [that there is no quantum collapse] is elegant, but heaven forfend it should be true! Sad to say, a reason to wish it false is no reason to believe it false. So, how many lives has Schrödinger’s cat?—If there are no collapses, life everlasting. But soon, life not at all worth living. That, and not the risk of sudden death, is the real reason to pity Schrödinger’s kitty. (2004, 21)

When I first presented this paper to a group that included both scientists and literary critics,7 a physicist in the audience observed that none of the people who endorse the quantum cosmology has had enough faith in the theory to attempt suicide. But rather than lacking faith, these people may have come to the same conclusion as Lewis: while survival is certain, so is the prospect of lasting forever in a state not worth living. Further, quantum suicide would not convince the skeptics of the validity of the many-worlds hypothesis, because in most worlds it would leave only a corpse. And in the few worlds where it would not, the survival of the suicide candidate could be attributed to luck, 8 not to necessity. The inevitable actualization of all possibilities opens another philosophical question, suggested in Niven’s story by the increase in crime since the advent of cross-world travel. What is the point in trying to act morally, if for every path I choose another me takes the opposite fork, if in some world Hitler behaves like Mother Teresa and Mother Teresa behaves like Hitler? Ethical 7. At a conference on Science, Text, Audience, Reception (STAR) at the Kavli Institute of Theoretical Physics, University of California, Santa Barbara, March 2005. 8. See also Bruce (2004), chapter 13 (“The Terror of Many-Worlds”), on the issue of quantum suicide and related problems.

Parallel W orlds  •  143

behavior presupposes free will, but if all possibilities must be realized at least once, I am predetermined to act one way or another, and my free will is an illusion: “No. There was no luck anywhere. Every decision was made both ways. For every wise choice you bled your heart out over, you had made all the other choices too. And so it went, all through history” (Niven 1971, 8). While I may think that my decision forces one of my counterparts to act the opposite way, my counterpart is under the same impression, and it is impossible to determine whether one of us acts freely. Choice can only exist in a multiverse where each possibility can be realized a random number of times, including zero. Then, by acting ethically, I increase both the total good of the multiverse and the quality of life in my own world, without dooming one of my counterparts to criminal action.

Conclusion In 1985, at a Nobel symposium devoted to possible worlds (proceedings published in 1989), the eminent Irish physicist John S. Bell outlined six possible approaches to the apparent contradictions of quantum mechanics. The most prominent are these: (1) Pragmaticism: be happy that the equations of quantum mechanics calculate probabilities accurately; don’t ask any philosophical questions. (2) Bohr’s complementarity: sometimes light behaves like a wave, sometimes it behaves like particles. (3) David Bohm’s idea that the behavior of subatomic particles is influenced by a hidden wave, so that instead of “particle” or “wave” we have “particles and wave.” (4) The many-worlds hypothesis. Bell, who endorses view 3, regards 1 and 3 as unromantic and 2 and 4 as romantic, because of the fascination they exercise on the imagination. By forbidding questions, pragmaticism kills any curiosity, while Bohm’s theory has attracted limited attention among philosophers and novelists. (Exceptions are Norris 2000 in philosophy and Goldstein 2000 in fiction). Complementarity, by contrast, is a favorite notion of postmodern theory, together with Einstein’s relativity, Heisenberg’s uncertainty, and Gödel’s incompleteness, even though these concepts are often interpreted in ways their original proponents would not approve of.9 As for many-worlds, its romanticism borders on sensationalism. The cover of Scientific American that advertised Tegmark’s article carried a headline worthy of a tabloid: “Infinite Earths in Parallel Universes Really 9. The postmodern recuperation concerns mainly the ideas of Einstein, who was a realist (cf. note 6), and of Gödel, who was a Platonist (Goldstein 2005), both positions rejected by postmodernism. Bohr’s and Heisenberg’s positions are much more compatible with the postmodern endorsement of constructivism and antirealism.

144  •  C hapter 7

Exist.” The imaginative appeal of the theory explains why in recent years the number of books about parallel universes in quantum mechanics has been out of proportion to the acceptance of the idea in the scientific community. The detractors of the many-worlds interpretation of quantum phenomena will say that it is an idea worthy of science fiction, but it is precisely its importance for one of the most vibrant genres of twentieth- and twenty-firstcentury literature that makes the theory fascinating for those interested in the workings of the narrative imagination.10 For literary scholars, who do not really care whether many-worlds cosmology is the best explanation of quantum mechanics, the interest of the theory is wide-ranging: it provides a powerful way to dramatize the what-if mode of thinking that makes us marvel at the major consequences of apparently insignificant events for our destinies (“if I hadn’t barely missed that plane I would be dead”); it offers new points of view on such fundamental questions as identity, ethical responsibility, and free will; it encourages questions regarding the nature of space and time; it rejuvenates the old theme of the double; and it creates intriguing narrative situations which would not be possible in a system of reality limited to one world.

10. For those interested in the motivation of plot from a technical perspective, the quantum cosmology raises the question of how narrative can support the idea of the incessant birth of an infinity of new worlds through a representation necessarily limited to a finite number of them. As David Bordwell observes (2002, 89), the branched narratives of film never follow more than a handful of paths, and the same holds (with probably higher limits) for literary texts.

CHAPTER 8

Impossible Worlds Dealing with Logical Contradiction

According to modal logic, a world is possible with respect to another world when it is linked to this world by a so-called accessibility relation (Kripke 1963).1 Since this relation can be conceived in many different ways, possibility and impossibility are relative concepts. Possibility could denote what may happen in the future, according to the laws of nature: for instance, the tower of Pisa could collapse some day and kill scores of tourists. In a more remote type of possible world, the laws of physics or biology no longer hold, so that princes can be turned into toads and vice versa. These worlds, which can be called unnatural, are widely represented in literature, as popular genres such as fairy tales, medieval fantasy, and science fiction demonstrate. Yet realistic texts that respect the laws of nature and fantastic texts that extend the possible beyond the natural do not exhaust the set of all possible texts: an important form of experimental literature creates worlds that cannot satisfy even the most liberal interpretation of possibility. These works transgress the logical laws of noncontradiction (not both p and ~p) and excluded middle (either p 1. Accessibility relations determine conditions of compossibility. They can be interpreted in various ways (Ryan 1991, ch.  2): identity of inventory (the two worlds contain the same objects), compatibility of inventory (the second world contains all the objects of the first, plus some native members), physical compatibility (the two worlds have the same natural laws), logical compatibility (the laws of logic operate in both worlds), and so forth. The ontological distance between the real world and a fictional world is a function of the number of accessibility relations that link them to each other. 145

146  •  C hapter 8

or ~p). They are much more challenging to the reader than works that simply describe the unnatural, because the mind cannot imagine p and ~p at the same time but can imagine the unnatural by relying in part on life experience, and in part on textual assertions. The surge of experiments with impossibility in literature and the visual arts coincides with developments in science that replace the Newtonian vision of a clockwork universe—predictable, knowable, and deterministic—with a world full of paradoxes: non-Euclidean geometry; elastic time and space that shrink and expand depending on the speed at which the observer travels; particles existing in a superposition of states and therefore satisfying both p and ~p; undecidable issues; entangled particles communicating without local connections; quantum tunneling; phenomena explainable only through mutually incompatible models (light behaving as both wave and particles); and either incompleteness or contradiction in mathematical systems. With modern physics, the natural and the impossible seem to exchange places: whereas in classical physics the impossible lies outside of nature, in modern physics, especially quantum mechanics, nature often seems to incorporate the logically impossible. By defying logic, literature and the visual arts rival the power of modern science to challenge the imagination. For convenience’s sake I call the semantic domain of the texts I have in mind impossible worlds, but as Umberto Eco (1979) has argued, the adequacy of the concept of world is questionable in these cases. Can a world be logically impossible and still remain a world? It all depends on what we understand by this term. The idea of textual, or fictional world is often used in a rather loose sense by literary scholars to refer to the content of the reader’s act of imagination. Since any text in a writing system that the reader can decipher evokes something to the mind, even if it consists of an invented language (as in Lewis Carroll’s poem “Jabberwocky”), it would take a totally unreadable collection of graphic marks to deprive a text of a world. In a narrower sense, the one that I will use here, the notion of textual world presents a referential dimension (as stressed in Stefanescu 2008): a textual world is a mental representation constructed on the basis of the propositions asserted in or implied by the text. The world of a text of narrative fiction can be thought of as a container filled with the characters and objects referred to by the text, as well as a network of relations between these entities. Reference, in this sense, does not require autonomous, actual existence: a fictional text refers to a world even though this world does not exist independently of the text, because the conventions of 2 fiction allow language to create objects by merely mentioning them. As in all 2. Fictional reference is what enables us to regard the expressions Madame Bovary, Emma, and Charles’s wife as describing the same individual.

I mpossible W orlds  •  147

uses of language, the mental construction of fictional worlds relies heavily on inferences. It would be not only unpractical to spell out all the facts that hold in a textual world; it would also be incredibly boring. This reliance of world construction on inferences calls, however, into question the adequacy of the notion of impossible world. Logicians believe that if a single contradiction penetrates a system of propositions, anything can be inferred, and every proposition and its negation becomes vacuously true (Goldstein 2005, 92). It would be impossible to imagine a textual world under these conditions, since textual worlds are constructed on the basis of a determinate set of propositions that the text presents as true. The texts I have in mind could be described as texts that cannot be true of any possible world, rather than as texts that refer to impossible worlds. Yet even if logic tells us that the phrase impossible world is an oxymoron, I retain it in this chapter because the readers of literary fiction have a broader of sense of worldness than logicians do, and because they do not treat inconsistencies as an excuse for giving up the attempt to make inferences. Literary works that project impossible worlds challenge readers to devise new strategies for making sense of texts, even if meaning does not arise from the vision of fully imaginable situations. Impossibility in a fictional world can take many forms, as the following sections demonstrate.

Contradictions The most obvious type of logical impossibility is a text that presents both p and ~p as facts in the fictional world. The contradiction can be as blatant as stating that “the cat is both dead and alive” (a technique typical of the folklore genre of nonsense poetry; Stewart 1978) or as imperceptible as hiding the conflicting statements within a dense paragraph, as does Robbe-Grillet when he begins a description with “outside it is raining” and ends it with “outside the sun is shining,” without suggesting a passing of time that could explain a change in weather; 1965b, 141). In this case only the most attentive reader will notice the breach of logic, unless the text is riddled with such contradictions. Contradiction in literature can present various degrees of granularity. On the largest scale, it opposes substantial segments of text. In The French Lieutenant’s Woman ([1969] 1980), by John Fowles, for instance, the last two chapters contain different endings: one in which the lovers, Charles and Sarah, commit to each other after a long separation, and one in which Sarah rejects Charles because she has found a fulfilling life without him. The two endings cannot be true at the same time, but within each of them the fictional world is perfectly consistent. The device does not ask the reader to construct an

148  •  C hapter 8

impossible world but rather to weigh the two endings against each other on the basis of such criteria as literary merit or consistency with the personalities of the characters. Next on the scale of granularity are contradictions that operate on the level of the relations between relatively short narrative segments. An example of this practice is Robert Coover’s “The Babysitter” (1969), a short story of 107 numbered paragraphs that cohere on the local but not on the global level. These paragraphs present different versions of what can happen when a couple goes to a party and leaves their children in the care of an attractive teenaged babysitter. In one version the babysitter is murdered, in another she is raped by her boyfriend and his buddy, in another the baby drowns in the tub, and in yet another the father leaves the party under the excuse that he needs to check on the children but really driven by the hope of having sex with the babysitter. The text tempts its readers to construct different stories by sorting out the paragraphs and ordering them in different sequence, but the last paragraph, which asserts as facts all the mutually exclusive developments, demonstrates the futility of such efforts: “‘Your children are murdered, your husband gone, a corpse in your bathtub, and your house is wrecked. I’m sorry. But what can I say?’ On the TV, the news is over, they’re selling aspirin. ‘Hell, I don’t know,’ she says. ‘Let’s see what’s on the late late movie’” (239). A third level of contradiction occurs when individual sentences, rather than entire narrative segments, clash with each other, producing what Brian McHale calls a “world under erasure” (1987, ch.  7): a world so thoroughly infused with ontological instability that readers cannot tell what exists and what does not. We have already seen this technique at work in the RobbeGrillet example mentioned above; an even more systematic use of sentencelevel contradiction occurs in Le Libera ([1968] 1978), by Robert Pinget, another representative of the French New Novel (a school that seems particularly fond of this technique). In Le Libera, a female narrator repeatedly contradicts the facts that she reports without apparently realizing the inconsistency of her tale. For instance, the same accident is said to kill a teacher, then to kill a certain student; it is, however, not this student who is killed but his brother; the brother is also found strangled in a field; a cook is first described as wonderful, then as a disaster; a school trip takes place in both the spring and the summer. While on the first three levels, contradiction involves a relationship between distinct units of text, on the last level it operates within the frame of the sentence itself. My example of this situation is a text of dubious narrativity, “Here We Aren’t, So Quickly,” by Jonathan Safran Foer (2010). The text consists of a list of mostly unrelated facts, evoked either in consecutive sentences (“He

I mpossible W orlds  •  149

was never happy unless held. I loved hammering things into walls”; 73) or in the constituent clauses of the same sentence (“You were not green-thumbed, but you were not content to be not content”; 72). The true originality of this text, compared with the other types of contradiction, lies in sentences that contain serious logical flaws: for instance, “I was always destroying my passport in the wash” (72 here and for subsequent quotations in this paragraph) denies the unique and punctual character of the act of destruction through an adverb (“always”) that presents it as either durative or iterable; “I was always struggling to be natural with my hands” is blatantly self-contradictory, since being natural is behaving without deliberate effort; “Everything else [beside the narrator and his wife being killed in a car accident] happened—why not the things that could have?” is a futile question, for if “everything else happened,” there is no point in asking why the things that could have happened did not: there are no such things—or at the most there is only one, the accident. The title of the story, “Here We Aren’t, So Quickly,” epitomizes the logical impossibility that permeates so many of its sentences: since here is a deictic referring to the speaker’s present position, it is incompatible with the negation of this position (aren’t); and since to be indicates a static, timeless position, it is incompatible with an adverb that suggests speed of movement through time (quickly). Foer’s nonsense sentences are not the product of a random juxtaposition of words; they are intricately wrought artifacts whose beauty lies in the elaborate analysis they require of the reader in order to diagnose the source of the weirdness. They make us logically and semantically smarter.

Ontological Impossibility In his groundbreaking book Postmodernist Fiction (1987, 10), Brian McHale identifies ontological concerns—that is, concerns with modes of existence—as the thematic dominant of the literature of the late twentieth century. A major form of this questioning is the creation of entities that belong simultaneously to incompatible ontological categories. This kind of impossibility is exemplified by the sentence “I am fictional.” The felicity conditions of this utterance could never be fulfilled, because the awareness of his own fictionality would attribute contradictory properties to the speaker: by saying “I am,” the character views himself as real, which means, as existing autonomously; but by recognizing himself as fictional, he acknowledges that he only exists through an author’s act of imagination. Or, to put this differently, I am is spoken from the perspective of the fictional world—within which the character exists indeed as a real person, if any kind of make-believe is to take place—while fictional

150  •  C hapter 8

reflects the perspective of the actual world, in which the speaker only exists as a literary character. The manifestations of ontological impossibility are known in narratology as metalepsis (Pier 2014, Genette 2004). Typical examples of this narrative device are readers or spectators penetrating the world of a fiction and intervening in its affairs (as in the movie Pleasantville, in which a teenager is transported into the world of a TV show and initiates its inhabitants into the lifestyle of the world he is coming from); and, vice versa, fictional characters breaking out of their world and invading reality (as in Julio Cortázar’s short story “Continuity of Parks” [1968] 1985, in which a reader is so totally immersed in a novel that the characters come to life and murder him). In these cases the breaking of boundaries is vertical, since characters from a lower, fictionally more real world penetrate a higher, imaginary world or vice versa, but metalepsis can also operate horizontally by importing characters from different literary texts and having them meet in the same world. This device is systematically exploited in the novels of Jasper Fforde, or in the comic book series The League of Extraordinary Gentlemen. (See introduction, note 3 for the cast of characters.) From the point of view of the actual reader, however, the boundary between the real and the fictional remains immune to ontological transgressions: it is only within a fiction that a reader can be murdered by a character in a novel, because real readers are located outside the system of fictionality. Metalepsis is not a breaking of boundaries between the (actually) real and the fictional but a breaking of boundaries that operates strictly within the diegetic levels of a fictional world. Its presence functions therefore as an obvious mark of fictionality. This self-referential, illusion-destroying effect explains why the device has become a dominant feature, some would say a trick of the trade of postmodern fiction.

Impossible Space Through the works of Magritte or Escher, impossible space has flourished in the visual arts; but its manifestation in literature is much more problematic because literature does not speak immediately to the senses and cannot therefore produce trompe l’oeil effects. In these paintings, impossibility derives from mutually contradictory properties: an object is both behind and in front of another; a staircase looks like it goes both up and down from a given point of view; the inside is outside and vice versa. It would be easy for language to create spatially impossible objects by juxtaposing mutually exclusive terms,

I mpossible W orlds  •  151

such as round square or flat sphere, but it is much more difficult to spin an interesting story that revolves around such entities. Moreover, impossible space should not be confused with unnatural space, that is, space that differs from our intuitive experience of the world. Consider these examples: 1. A world with only two dimensions or one dimension, such as the Flatland and Lineland described in Edwin Abbot’s Flatland: A Romance of Many Dimensions. Such worlds are certainly different from our world, but they are easy to imagine, since all it takes to visualize them is to flatten our representation of reality. 2. A world made of more than three dimensions. Such a world can be mathematically described (string theory postulates a universe of eight to ten dimensions), but it is impossible to represent for the imagination, if by represent one means mentally picturing. In Abbot’s Flatland, when the hero A. Square asks the inhabitants of a three-dimensional world to take him to a four-dimensional one, they reply that there is no such world. 3. A world that scrambles real-world geography. Jan Alber, in his discussion of unnatural spaces, mentions a short story by Guy Davenport, “The Haile Selassie Funeral Train,” “that turns Europe into an unnatural collage-like zone: from Deauville in Normandy the train passes though Barcelona along the Dalmatian coast, through Genoa, Madrid, Odessa, Atlanta (Georgia, U.S.A.) and back to Deauville” (2016, 199). According to possible worlds theory, there is no reason why there could not be a world where the counterparts of real-world cities are distributed in the way suggested in the story. 4. A world with very strange objects, such as the floating island of Laputa in the realm of Balnibarbi in Gulliver’s Travels that defies the laws of gravity. Of this island Swift writes: “The Flying or Floating Island is exactly circular, its Diameter 7837 Yards, or about Four Miles and a half, and consequently contains ten Thousand Acres. It is three Hundred Yards thick” ([1726] 2001, 155). 5. A wormhole universe, allowing one to pass into another universe through black holes, rabbit holes, or secret passages in the back of closets. 6. A universe with parallel worlds, as in the stories discussed in chapter 7.

In her book Narrative Space and Time: Representing Impossible Topologies in Literature, Elana Gomel defines narrative manifestations of impossible space as “textual topologies that defy the Newtonian-Euclidean paradigm of homogeneous, uniform, three-dimensional spatiality” (2014, 3; italics original).

152  •  C hapter 8

All of the above examples transgress this paradigm in one way or another. But if physicists have developed cosmological models that transcend Euclidean geometry and Newtonian physics, this means that the seemingly impossible spaces that correspond to these more comprehensive models can be real after all, since it is the business of physics to describe reality. Gomel admits so much when she writes, “impossible spaces are unrealistic because realism is underwritten by the Newtonian paradigm. But they are not necessarily unreal.” And she adds, quoting literary theorist Susan Strehle (1992), who has studied the influence of quantum mechanics on fiction, “today realism itself has become impossible” (30). I would not go as far as declaring realism obsolete, because realism concerns the world of everyday experience, and the Newtonian-Euclidean model remains valid within this sphere. In fact, if we can imagine worlds of types 5 and 6 above, it is because we can expand our innate Newtonian-Euclidean models of reality to planetary systems and to universes including many worlds. But if one accepts that the wildest speculations of physicists may be accurate representations of the real, what does it take for space to be impossible? In chapter 7 we saw that physicist Max Tegmark imagines the possibility of a world “that is simply an empty dodecahedron” (2003, 49). Yet if the universe is inherently mathematical, as Tegmark believes, this dodecahedron world would still be subject to the laws of mathematical logic. The impossible spaces I discuss here are spaces that transgress the laws of noncontradiction and excluded middle; but how can such constructs be made to contribute to narrative meaning? As examples 1 through 6 show, it is difficult to tell whether a space is impossible or simply strange. The island of Laputa is strange, but it would be possible if it weren’t for its measurements, which make the inside larger than the outside—a mathematical contradiction.3 In the movie Pleasantville, in which the eponymous town functions as a caricature of the world of the TV shows of the fifties, the town is built along two roads: Main Street and Elm Street. When people reach the end of Main Street, they find themselves back at the beginning.4 This could be explained by a curved topography whose 3. Actually, the whole world of Gulliver’s travel is mathematically impossible given the measurements (in latitude and longitude) given by Gulliver. As I write elsewhere (“Maps and Narrative,” ch. 3, in Ryan, Foote, and Azaryahu 2016, 53), using research by John Robert Moore (1941), “Luggnagg is located at 29 degrees N. and 140 degrees E., and Balnirabi [sic] at 46 degrees N. and 177 degrees W. Yet the two islands are only separated by 150 miles—something impossible on a globe the size of the Earth. Moore concludes that Luggnagg is both northwest and southwest of Balnirabi.” There are numerous examples of such impossibilities in the book. Alber would read them as “satire” (cf. note 7). 4. This fact is told to the heroine by a professor who lives in Pleasantville, but it is not implemented visually in the film.

I mpossible W orlds  •  153

dimensions curl back upon themselves (as in the video game Pac-Man:5 if you exit on the right of the screen, you re-enter on the left because space is circular). But since the inhabitants of Pleasantville experience Main Street as a normal space occupying a segment of an infinite dimension, Main Street represents two incompatible topologies: a curved world and a flat world. This impossible topology symbolizes the self-enclosed character of the community and its blissful ignorance of the problems of the larger world. Another example of impossibility built from a possible structure is the case of the Moebius strip, where the inside becomes the outside and the outside the inside. Moebius strips can be constructed by twisting a strip of paper, and a character in a realistic novel could easily make one. But in a story discussed by Patricia García, “Dejen Salir” [Exit], by José Ferrer Bernejo (1982), the hero is caught in a Madrid metro station shaped like a Moebius strip, and he cannot find his way outside (García 2015, 121–25). Here the Moebius strip is not just an isolated object in a basically Euclidean-Newtonian world; it becomes the world itself, and exiting the metro station, an easy task in a normal world, becomes an impossible pursuit because this world has no outside. While in “Dejen Salir” impossibility arises when a possible object becomes a whole world, in Mark Z. Danielewski’s House of Leaves (2000) an impossible object is contained within an otherwise normal topology. The impossible object is a house that is larger on the inside than on the outside. Though the measured difference remains constant—about an inch—the inside expands into a hallway and then into a labyrinth of seemingly infinite dimensions. An expedition is sent to measure this labyrinth and to create a cinematic record of its configuration, but the exit is never found, and many of the explorers disappear or become insane. Through this expansion, the impossible object contaminates the entire world. The structure of the house is replicated on the level of the book as a material object through an outside—the cover—visibly shorter than the pages of the inside. Impossibility runs, however, deeper than space in the world of House of Leaves. The story of the house, known as the Navidson record, belongs to a novel (or documentary?) allegedly written by a blind man, Zampanò, and 5. For a discussion of impossible space (and time) in digital fiction, especially in video games, see Ensslin and Bell (2021). However, some of the game spaces the authors describe as impossible are merely strange: for instance, the game space of Inkubus (2014), in which the user explores a human body from the inside, does not necessarily violate the laws of logic or of Euclidean geometry. A more convincing example of impossible space is the grandmother’s house in the 2009 game The Path (an adaptation of “Little Red Riding Hood”), which is clearly bigger inside than outside, like the house in House of Leaves discussed in this chapter. It looks like a little cottage, but the player must move through endless rooms and hallways to reach the grandmother and the end of the game.

154  •  C hapter 8

found after his death by a character named Johnny Truant, who edits the manuscript and adds footnotes to Zampanò’s text. Yet another editor-character packages Zampanò’s narrative, Truant’s text, and various other documents— such as letters sent to Truant by his mother, Pelafina—into a book, adding his own notes to Truant’s comments. In a breach of probability that does not quite reach the logically impossible, Zampanò’s blindness and solitary life make him the most unlikely author for a manuscript that focuses heavily on visual media. Ontological boundaries are violated when Pelafina, afraid that her letters are being intercepted, asks Truant to put a checkmark in his next letter to demonstrate that he has received hers; but the checkmark appears in Zampanò’s text, and Zampanò does not know anything of Pelafina’s (or Truant’s) existence. The novel also dismantles the physical space of the text through a wild play with typographical presentation that confronts the reader with constant decisions: should she read first Zampanò’s narrative about the impossible house and then Truant’s notes, or should she read them concurrently; should she read the text that has been crossed out, or should she skip it; should she read the medallions of texts shown on some pages before the text that frames them, or the other way round? House of Leaves is a narrative presented in book form, but its graphic design subverts the sequential reading protocols traditionally mandated by both narrative and books. The difference between the inside and the outside of the house is not only a wormhole that leads into an alternate world of darkness and horror; it is also the initial inconsistency that, according to logic, opens the floodgates to all kinds of paradoxes.

Impossible Time For most people, the first idea that comes to mind as an example of impossible time is time travel. Though time travel can lead to logical contradictions, such as the grandfather paradox mentioned in chapter 7, I will not discuss it here, because its impossibility is far from certain. Einstein’s special theory of relativity from 1905 suggested the possibility of travel into the past, while the mathematician Kurt Gödel pondered the possibility of travel into the future (Nahin 1999, xviii). And indeed, if time is conceived as the fourth dimension of a totality called spacetime, travel through time poses few problems to the imagination, since we can move across it as we do across space. Here I concentrate on more counterintuitive treatments of time. Time is a much more abstract, much less graspable concept, than space; yet while we cannot capture its nature in words, as St. Augustine famously observed (“What then is time? If no one asks me, I know what it is. If I wish to

I mpossible W orlds  •  155

explain to him who asks, I do not know”),6 we have reasonably firm intuitive beliefs about its properties. It is the contradiction of these beliefs that leads to temporal impossibilities. Our most fundamental intuition about time tells us that it flows in a fixed direction,7 though this direction is a matter of debate: according to one conceptual scheme, time flows from the future to the past, since future moments become present and then past; in another scheme, time flows from the past to the future, since the future is ahead of us and we are marching toward it. The axiom of the fixed directionality of time can be broken by reversing its flow. Two novels that attempt this conceptual tour de force are Counter-Clock World, by Philip K. Dick (1961), and Time’s Arrow, by Martin Amis (1991). Yet if the head of the arrow of time can stand for either the past or the future, depending on the particular conceptual scheme, how can one distinguish the future from the past, and how can one distinguish “normal” from “reversed” time? It takes an external point of reference to determine in which direction time is flowing; in both novels, this reference is provided by familiar biological processes and social scripts: characters die before they are born, eating consists of regurgitating food onto a plate, and healthy people get sick after a visit to the doctor. In Time’s Arrow, in addition, the reversal of time is suggested by a sequence of historical events familiar to the reader: the narrator is the “soul” of a Nazi doctor who died after emigrating to the US, and he relives his alter ego’s life in reverse order, from cold war America to World War II, and from the liberation (or rather, creation) of Auschwitz to the rise (or rather, decline) of Nazism in Germany. But the reversal of time can be conceived along arrows other than biology or chronology: one of them is causal (since causes precede effects in normal time, they should follow them in reversed time); the other is cognitive (we know what lies in the past, but we don’t know what lies in the future). If novels were fully consistent in reversing the flow of time, they would have to invert the causal and cognitive arrows, but this reversal would 6. Quoted from Morris (1984, 8). 7. As Rebecca Goldstein writes, however, “in relativity theory, for example, time doesn’t flow, but rather is, as the fourth dimension, as static as space” (2005, 42). Moreover, the equations that describe the laws of nature do not generally support the idea of an intrinsic directionality of time. From a physicist’s point of view, most processes are time-symmetric, in that they would unfold in the same way if the direction of time were reversed. But there is a notorious exception within physics to this statement: the second law of thermodynamics, which states that the entropy of a closed system is steadily increasing—in other words, that the closed system of the universe is evolving from order to chaos. A state of greater entropy can be called the future with respect to a state of lesser entropy. In Time’s Arrows: Scientific Attitudes Toward Time, Richard Morris (1984, 131–48) mentions four other arrows of time: the expansion of the universe, the decay of K-meons, the directionality of electromagnetic waves, and what he calls the “psychological arrow”—our intuitive sense that we are moving toward the future.

156  •  C hapter 8

deprive characters of any form of free will and agency, since the flow of time would inexorably carry them toward a state that is already determined. This in turn would make planning and consequently plot pointless, since the purpose of plans is to control our destiny. To preserve narrative tension, time-reversed narratives typically limit their reversal to history and biology (and even the reversal of history is controversial because the causality that explains historical events works only in one direction: you can explain World War II through the rise of Nazism in Germany in the 1930s but not vice versa, unless you change the facts). In Dick’s novel the characters remain unaware of what lies ahead of them, and they make plans to affect the future as if they lived in normal time; whereas in Amis’s novel the narrator is the only character who experiences time backwards: what is for him an unknown future is a known past for the others, and what is shared history for the others is for him a future mysterious yet unavoidable. Deprived of the freedom to create his own destiny, he has no choice but to discover passively the life that his alter ego the Nazi doctor has already written for him. Another fundamental belief about time tells us that the future is open whereas the past is written once and for all: you can affect the future though your actions, but you cannot undo the past. Since philosophers cannot demonstrate that changing the past leads to contradiction, the impossibility of doing so must be regarded as an axiom. Those who have attempted to explain their intuitive feeling that the past is written once and for all have often fallen into redundancies and tautologies. An especially blatant case is this passage from a college textbook by John Hospers: “This is an unchangeable fact: you cannot change the past. That is the crucial point: the past is what happened, and you can’t make what has happened not have happened. Not all the king’s horses or all the king’s men could make what has happened not have happened, for this is a logical impossibility” (quoted in Nahin 1999, 259). In La Moustache ([1986] 2005), Emmanuel Carrère explores the trauma that would arise if the past could be changed for one person but remained stable for all others. A tragic chain of events is set in motion when the narrator and main character innocuously decides to shave off the moustache he has been wearing for ten years to surprise his wife. But when his wife comes home she shows no surprise at all. The narrator suspects that she is playing a trick on him, but the next day at work his colleagues also behave as if nothing had changed. This is only the beginning of an inexorable process that disintegrates the narrator’s personal history piece by piece and replaces it with another life. For the greater part of the novel, the events whose occurrence is denied by the other characters precede the beginning of the story, but in the last chapter,

I mpossible W orlds  •  157

just before the narrator is driven to suicide by the loss of his memories, which means the loss of his identity, the past being erased concerns the events of the earlier chapters. The novel tells us how the narrator flees to Hong Kong and then to Macao to escape from the people who deny his past; but when he arrives in Macao he finds his wife in the hotel room, enjoying an ordinary vacation. This means that the events represented in the earlier chapters never took place. At this point the novel becomes a self-destructing artifact that denies what is generally considered the main function of narrative: its ability to tell about and to preserve the past.

Impossible Texts It may seem a matter of common sense to treat impossible texts according to Ludwig Wittgenstein’s recommendation: “Whereof one cannot speak, one should remain silent” (1981, 7). Since these works do not exist, why should we bother with them? But at least some impossible texts can be imagined, and therefore described in language. Here a caveat is in order. An impossible text is not a text that describes impossible objects and worlds; otherwise, all the texts discussed in this chapter would fall into this category: it is a text that cannot be produced as a physical object, regardless of content. Inventing texts so paradoxical that they can only exist in the imaginary mode is one of the favorite pursuits of Jorge Luis Borges. Most of these fictional fictions involve a form of infinity. The most celebrated of them is the novel “The Garden of Forking Paths,” written by the Chinese author Ts’ui Pen and mentioned in the eponymous short story. (See chapter 7 for a fuller discussion.) In Ts’ui Pen’s novel, all the outcomes in fact occur; each is the starting point for further bifurcations. Once in a while, the paths of that labyrinth converge: for example, you come to this house, but in one of the possible pasts you are my enemy, in another, my friend. (Borges 1999, 125)

It is not because of the spatialization of time discussed in chapter 7 nor because of the contradictions created by the merging of paths (a character being both a friend and an enemy to the narrator) that Ts’ui Pen’s novel is an impossible text, but simply because of its ambition to capture possibility in its totality, a task that would require infinite time. Another of Borges’s impossible creations is the Book of Sand, also mentioned in an eponymous story. This book has neither beginning nor end:

158  •  C hapter 8

wherever one opens it, there are always some pages between the cover and the current page, and when one turns a page, one lands at any distance from the previous page, so that a complete and sequential reading becomes impossible. While in “The Book of Sand” infinity concerns the book as a physical object, in The Thousand and One Nights it affects the act of narration. On the six hundredth second night, Borges writes in his essay “Partial Magic in the Quixote” (1964), Scheherazade tells the Sultan his own story (a fact that I could not verify by looking at a complete version of The Thousand and One Nights). This leads to infinite recursion, since this story contains all the stories that Scheherazade tells the Sultan to postpone her execution, including the story of the six hundredth and second night.

Impossible Worlds: A Challenge to the Reader What is the point of creating fictions with impossible worlds? Another way to formulate this question is to ask, with Jan Alber (2016), what should readers do with these worlds? The mental strategies for making sense of impossible worlds must be measured against the standard procedure by which readers construct standard fictional worlds, whether realistic or fantastic. Relying on David Lewis’s analysis of truth in fiction (1978), I have called this procedure the principle of minimal departure (Ryan 1991); Walton (1990, 144–50) calls it the reality principle. This principle enjoins readers to construct fictional worlds as the closest possible to their model of reality, amending this model only when it is overruled by the text. This means that readers should not make gratuitous changes. Thomas Pavel has suggested that in the case of many texts, especially those with impossible worlds, readers expect a “maximal departure” (1986a, 93). This term may very well describe the reader’s general expectations with certain types of text (for instance, nonsense poetry), but, unlike minimal departure, such a principle offers no guidance for making inferences and filling in the blanks in the text. Imagine a fiction that mentions a restaurant: minimal departure activates standard scripts about this kind of institution, but what would we imagine under maximal departure? That the restaurant is really a pigpen? That it is a prime number? Or that it pays its customers to eat there? Since concrete objects do not have specific opposites, maximal departure comes up empty in the construction of impossible worlds. How, then, do we make sense of these worlds? There is no single strategy that works for all the types I have described but rather a variety of possible operations whose relevance depends on the particular text. Some of these

I mpossible W orlds  •  159

interpretations can be called naturalizing because they aim to preserve the logical integrity of the fictional world:8 • Mentalism: The inconsistencies can be explained away as dream, hallucination, or the dementia of an essentially unreliable narrator. If the story is a dream or hallucination, it must stage a return to a normal world (as in Alice in Wonderland); if the contradictions are due to the narrator’s mental confusion, the reader must be able to reconstruct the normal world that this discourse misrepresents. • Figural interpretation: The inconsistencies do not correspond to facts; they are only ways of describing certain phenomena. For instance, if Sherlock Holmes is reported to have squared the circle, this could mean that he can solve any problem. In House of Leaves, the fact that the inside (=emotions) is bigger than the outside could be interpreted as the triumph of love (between Karen and Navidson) over material circumstances (Alber 2016, 190–91). • Many-worlds and virtualization: The mutually incompatible elements are part not of the same world but of different possible worlds. This idea could be used to naturalize branching stories, such as The French Lieutenant’s Woman. From the point of view of the world of each of the two endings, this world corresponds to the fictional facts, and the other ending is a nonactualized possibility, a purely virtual scenario. The texts that lend themselves to this operation can be described as polycentered because they ask of the reader to recenter herself into different worlds.

When naturalizing explanations fail, the reader must accept contradiction as integral to the fictional world. In this case there are two possibilities: 8. Jan Alber comes up with a related, but partially different, list of “strategies for making sense of the unnatural” (2016, 47–48). Since my concern is the impossible, not the unnatural, my list is shorter. Alber’s list is as follows (explanations in parentheses are Alber’s; those in square brackets are mine):

1. 2. 3. 4. 5. 6. 7. 8. 9.

The blending of frames [= imagining hybrid entities] Generification (evoking generic conventions from literary theory) Subjectification (reading as internal state) Foregrounding the thematic [as opposed to the mimetic] Reading allegorically Satirizing and parody Positing a transcendental realm [such as hell, purgatory, or paradise] Do it yourself (using the text as construction kit to build your own stories) The Zen way of reading [= an aesthetic experience that foregoes logic]

160  •  C hapter 8

• Dreamlike reality: In contrast to the mentalist interpretations that limit contradiction to the alternate reality of a dream world, this interpretation endows the actual world of the textual universe with the characteristics of dreams: fluid images, objects undergoing incessant metamorphoses, and general lack of ontological stability. A prime example of a narrative with a dreamlike reality is Franz Kafka’s “A Country Doctor.” It is in this category that I place Robbe-Grillet’s In the Labyrinth and Pinget’s Le Libera, though Le Libera could also be interpreted as a case of mentalism. • Swiss-cheese world: Whereas in dreamlike realities the entire fictional world is filled with contradictions, in this case the irrational is contained in delimited areas that pierce the texture of the fictional world like the holes of a Swiss cheese. The reader can still use the principle of minimal departure to make inferences about the fictional world, as long as these inferences concern the solid areas. This is how I personally interpret narratives involving ontological, spatial, and temporal paradoxes. For instance, the hero of La Moustache is the only character in the novel whose past is constantly changing; the house in House of Leaves is the only one that functions as a portal into a terrifying world. By confronting a normal world with an irrational one, the Swiss-cheese configuration makes the experience of the irrational much more dramatic than in dreamlike realities because the experience of the protagonist clashes with the normal world in which other characters seem to live comfortably.

Rather than preserving the idea of an autonomous storyworld—and therefore maintaining the possibility of an immersive experience, for immersion requires a recentering of the imagination to an alternate world—interpretive strategies can emphasize the narrative process: • Metatextualism and do it yourself: The conflicting versions are not alternative worlds but alternative drafts of a novel in progress. This strategy could be applied to “The Babysitter” and The French Lieutenant’s Woman. In the case of “The Babysitter,” the reader is offered a construction kit to make her own stories. • Fragmentation: The text is like a collection of pieces from several different and incomplete jigsaw puzzles. Readers assign the contradictory fragments to different worlds, but none of these worlds can be fully constructed, and they prevent recentering. They can only be contemplated from an external, metatextual perspective. Robbe-Grillet’s In the Labyrinth and Pinget’s Le Libera could be examples of this way of reading.

I mpossible W orlds  •  161

All texts with impossible worlds have the effect of drawing the reader’s attention to the textual origin of the storyworld. I believe, however, that the only people who can be satisfied with a purely metatextualist interpretation are literary theorists. Most readers will do whatever they can to salvage a world in which they can achieve some degree of immersion, because makebelieve, the mental game responsible for this experience, corresponds to a fundamental need of the human mind and is simply more pleasurable than the game of pure self-reflexivity. It would be easy to say of texts with impossible worlds that “this is just words.” It would be just as easy to ignore the contradictions and to build a coherent world. But impossible worlds cannot be fully contemplated from either of these two stances. If the texts that construct such worlds pose a challenge to the reader, it is because they require the ability to shift back and forth between a narrativist-illusionist and a textualist perspective, so as to evaluate the text both as a (questionable) representation of life experience and as a virtuoso verbal performance that pushes back the limits of the textually (im)possible.

CHAPTER 9

Virtual Worlds Narrativity and VR Technology

When the great leap forward of computer technology reached the general public in the 1980s and ’90s, there was a pervasive sense that we were witnessing the birth of a new medium of representation, a medium that would make as significant a difference for art and entertainment as the great inventions of the past, such as writing, film, and television. One of the forms this new medium was expected to take was virtual reality (VR), a technology promoted at the time as the “next big thing” that the computer would bring into our lives. In this chapter, I propose to explore the potential of VR technology as a narrative medium. As a preliminary, I briefly sketch the history of the VR idea and place it in the context of the various forms of digital entertainment.

VR: The First Wave of Publicity VR was conceived by its advocates as an “immersive, interactive experience generated by a computer” (Pimentel and Texeira 1993, 11), and the key notion was that of presence. Participants would experience an image created by a computer as if it were “the real thing,” this is to say, as if it were a material reality. They would be able to modify the simulated environment through the gestures of their own body or through voice commands. The notion of pres-

162

V irtual W orlds  •  163

ence means that the computer should disappear from active consciousness and be replaced by the object of the simulation. In 1997 Janet Murray’s popular book Hamlet on the Holodeck explored the possibility of turning narrative into the kind of experience that defines VR. The holodeck is a fictional technology from the TV series Star Trek. It consists of an empty cube on which a computer projects a three-dimensional world simulation. The user steps into this world and interacts with synthetic characters operated by artificial intelligence (AI). No matter what the user says or does, the system responds coherently and integrates the user’s input into a narrative arc that sustains interest. The user thus becomes a character and experiences the narrative from a first-person perspective. If the holodeck represents a valid model for digital narrative, this means that the computer does not have to rely on a fixed story, as do most computer games. A coherent plot must emerge from the live interaction between the human participant and the computer-generated environment. Narrative is about the actions of people, about their relations to other people and to their environment, but not every sequence of actions and events constitutes a wellformed narrative. As Aristotle observed in the Poetics (5.1; 1996, 13–14), stories must have a beginning, middle, and end; they must also represent some kind of conflict that comes to a resolution. In regular stories the author has complete control over the plot, but in interactive stories the plot must emerge from the interaction between the user and the computer. This means that the story must be generated partly by the user and partly by artificial intelligence. A lot of effort has been devoted since the year 2000 to develop a so-called narrative intelligence (Mateas and Sengers 2003), but the results have been meager. I do not know of any computer-generated narrative worth reading for its own merit. As I argue in Ryan 2015, if computers still cannot generate entertaining, noninteractive stories when the machine has total control over the storyworld, the difficulty of the task is taken to a much higher power when the story must integrate the unpredictable input of the user. In the holodeck, for whatever the user says or does, the computer must understand the action of the user, respond in an appropriate way, and steer the storyworld toward an appropriate ending that brings narrative satisfaction to the human participant. The dynamic generation of interactive stories is not the kind of problem that could be solved in the future by a more advanced AI, because it is not a technological but a logical and psychological problem. The AI researchers Ruth Aylett and Sandy Louchard describe this problem as “the interactive paradox”: “On one hand the author [in this case the computer] seeks control over the direc-

164  •  C hapter 9

tion of narrative in order to give it a satisfactory structure. On the other hand a participating user demands the autonomy to act and react without explicit authorial constraint” (2004, 25). There is no solution to the paradox but only acceptable compromises: either restrict the freedom of the user in order to create a coherent narrative, or expand the freedom of the user and sacrifice narrative form. As the year 2000 neared, it became evident that neither the vision of VR developers nor holodeck narratives were going to materialize any time soon. Jaron Lanier, one of the pioneers of VR, told an interviewer: [As for the waning of virtual reality from public attention,] “I bear some of the blame for it. I always talked about virtual reality in its ultimate implementation and when that didn’t happen, interest declined. Because everyone wanted the holodeck from Star Trek, virtual reality couldn’t fulfill its promise so quickly” (quoted in Ditlea 1998). After 2000 VR continued to be developed in the scientific sectors, and it found a number of practical applications (Slater and Sanchez-Vives 2016). But the public expected new forms of entertainment, and when they did not materialize, VR disappeared from the radar of the media.

The Two Poles of Digital Entertainment Instead of Hamlet on the holodeck, we got two types of digital entertainment. At one pole is the esoteric domain of electronic literature, represented by hypertext fiction as well as by three (to this date) collections of texts on the internet gathered by the Electronic Literature Organization. Hypertext fiction enjoyed for a while a cult following in academia, and it was touted as the future of the novel, but it never caught on with the general public, because by breaking up stories into fragments, and by giving the reader a choice of links to follow, it implemented interactivity at the cost of immersion. Far from turning readers into creative co-authors, as its early advocates claimed (Bolter 1991; Landow 1997), hypertext deprived them of such basic narrative effects as surprise and suspense (Sternberg 1992), which depend on a precise authorial control of the release of information. Such control is impossible when readers can choose multiple paths through a network of interconnected fragments. As for the texts gathered in the ELO collections, most are dominated by a postmodern aesthetics that locates literary value in self-reflexivity and attention to the medium rather than in world construction. These texts make the computer visible through the exploitation of dysfunctionality, such as deliberate glitches, vanishing inscriptions, nonsense generation, pointless interaction, parodies

V irtual W orlds  •  165

of digital tools, and heterogeneous collages of data randomly fished from the internet (Emerson 2014). As a general rule, electronic literature endorses the interactive dimension of VR, but it rejects its immersive ideal because it regards it as incompatible with its ideal of self-reflexivity. Yet, according to VR researchers Mel Slater and Maria Sanchez-Vives, illusion in VR is never total. They argue that users only experience presence when they remain aware that the object of their perception is a computer-generated image. In real life, we take the presence of the environment that surrounds us for granted and we do not reflect on it; in VR, by contrast, the experience of presence should become a cause of wonder and a potential source of pleasure. As Janet Murray writes: “It is in fact this double consciousness that makes VR so thrilling—our sense that the virtual world seems so real despite our knowledge that our feet are still planted in this world” (2020, 19). At the other pole of digital entertainment is the wildly popular genre of computer games. Judging by their addictive nature, these games fully realize the VR ideal of an immersive, interactive experience generated by a computer. While in hypertext fiction interactivity came at the cost of immersion, in games interactivity has an immersive effect, since players are consumed by the desire to solve problems. The immersivity of computer games comes not only from the agency given to players and from the desire to beat the game or other players; it also comes from the construction of sensorially rich gameworlds, which players can explore and where they feel at home. In contrast to traditional board games or sports games, such as chess or football, computer games place players not in an abstract space but in a concrete environment full of recognizable objects and characters. The tasks given to players are not actions that receive meaning only by the rules of the game, such as kicking a ball into a net; they are instead inherently meaningful and beneficial actions, such as killing dragons, rescuing princesses, or escaping enemies. There are few princesses and no dragons in the real world, but if dragons existed, and if the situation presented itself, surely most people would jump at the opportunity to save a beautiful and rich princess from an evil dragon. This concrete, representational dimension creates a close connection between computer games and narrative, for if games construct worlds, and if players can perform actions that change the state of these worlds, there must be some kind of story that unfolds in the gameworld, thanks to the player’s activity. But even though they are both interactive and immersive, computer games are not the holodeck, because they subordinate narrative to gameplay, and, as a result, their underlying narratives are not the focus of attention but a means toward a goal—the goal of motivating a player to play. This subordination to gameplay explains why there is limited variety in game narratives compared

166  •  C hapter 9

with the amazing diversity of book or film narratives. The most common narrative pattern in computer games is an adaptation of the archetype of the hero’s journey, as defined by Joseph Campbell (2008), Vladimir Propp (1968), and others: a solitary hero accepts a mission, goes on an adventure, passes a number of tests, and returns home victorious (or sometimes defeated, as may be the case in computer games). Standard game narratives follow predefined scripts that force players to pass though the same sequence of checkpoints, but each player creates variations on the fixed pattern, since no two players will perform the exact same actions. Branching structures can produce different stories, especially stories with different endings, but even though the player’s actions determine how the story will end, the various possible endings are built into the system rather than created on the fly. With electronic literature and shooter-type or adventure computer games, digital entertainment has reached the two poles of the cultural spectrum— the pole of experimental, avant-garde literary forms, frequented by a rarefied, mostly academic public; and the pole of popular culture, frequented by the masses. But between these two zones, which I have called the North Pole and the Tropics (Ryan 2007a), lies a third area, which I call the Temperate Zone, and which is frequented by an educated but critical public, a public that desires neither the deconstruction of narrative nor its reduction to stereotyped forms and its subordination to gameplay. All significant narrative media have conquered the Temperate Zone, but this conquest has been particularly problematic for digital technology. Independent games have been working to fill this gap, especially through the development of a game genre known as environmental storytelling or walking simulator, and represented by such games as Dear Esther or What Remains of Edith Finch. This genre follows a pattern that I call epistemological, the pattern of the detective story. The player explores a computer-simulated world and uncovers a story that took place in this world, often by finding telltale objects such as letters and diaries, or simply by listening to voices that narrate past events when the player reaches certain locations in the game world. The pleasure of these games does not lie in an original or exciting gameplay—the player’s activity is limited to moving through the gameworld and looking at objects—but resides in curiosity for the story, in the satisfaction of extracting it from the environment, and above all in the visual appeal of the gameworld. Yet, in a polemical article titled “Video Games Are Better Without Stories” (2017), the prominent game theorist and game developer Ian Bogost questions the value of the narrative experience provided by environmental games. He argues that the story told in What Remains of Edith Finch would be better served by making it into a movie or telling it in a novel. And indeed, if a story

V irtual W orlds  •  167

truly captivates the reader, the effort needed to discover it by finding a way to progress through the gameworld may be more annoying than gratifying. Movement through the gameworld is restricted by an invisible code, and the player must find the proper paths or points of access to uncover new information. But while the game genre of environmental storytelling may frustrate people who play for the plot exclusively, I cannot agree with Bogost that video games are better without stories. As already mentioned, these games owe much of their success to their ability to give narrative meaning to the player’s actions; without narrative structure, we would still be playing Tetris or tic-tac-toe. On the other hand, I tend to agree with Bogost that game stories have not reached the artistic heights and diversity of the best novels, dramas, and films, and possibly never will. But even if their narrative potential does not rival that of novels or film, computer games still have much to gain by trying to realize this potential.

The Rebirth of the VR Dream The three genres I have discussed so far—electronic literature, standard games focused on gameplay, and independent games focused on the discovery of a story—were pretty much the state of the art of digital textuality in the mid2010s. But then VR made a surprising comeback. The trigger of its rebirth was the development of new hardware to replace the clumsy, heavy equipment of the 1990s. In 2014 Mark Zuckerberg, the founder and CEO of Facebook, purchased Oculus Rift, the maker of a relatively cheap and lightweight VR helmet, for two billion dollars. The supposedly imminent second coming of VR spurred renewed talk about the narrative potential of the medium, and a few how-to books about VR narrative appeared (Riggs 2019, Tricart 2018, Bucher 2017, Bosworth and Sarah 2018). But where are the applications, where is the content? In the remainder of this chapter, I discuss some of the existing narrative applications of VR in order to assess the storytelling potential of the medium. But as a preliminary, let me return to the experience of immersion.

Types of Immersion When we talk about immersive, interactive experiences, it is clear that interactivity is a property of the medium or, if you prefer, of the system. We can safely say that computer games are interactive but that books and standard comics

168  •  C hapter 9

and films are not. But in contrast to interactivity, immersion is a mental experience, and its relationship to the medium is much more problematic. Most of the books I have read about VR and narrative speak of immersive media, as if some media had immersivity built into them through their technological features. According to this view, media can be ranked on a scale of immersivity which would put text at the bottom, followed by still pictures, comics, film, computer games, and VR. The position of a medium on the scale would depend on how many different senses it affects, and on how much agency it grants to the user, for interactivity is an important factor of immersivity. In another conception of the term, immersion is an experience, an imaginative state, that can be triggered by any medium; you can be “lost in a book” as well as immersed in a computer game. Immersion, in this sense, is the user’s response to the author’s design; and making it happen is a strength of some works. You can have immersive books and nonimmersive VR projects. Here I will settle for a compromise between the two conceptions: I agree that media differ from each other through their immersive potential, but this potential may or may not be actualized in a specific work. As a mental state, immersion can be divided into several categories. First, there is ludic immersion—the passion a player brings to playing a game, solving problems, and beating opponents. This type of immersion presupposes an interactive environment and is independent of any kind of concrete world and narrative content; it is experienced by chess and bridge players as intensely as by players of computer games such as World of Warcraft or Grand Theft Auto. Then there is what I will call mimetic immersion, or narrative immersion, which is a response to the concrete representation of a world that evolves in time. This type of immersion can be provided by a variety of media: by a literary narrative, by film and TV, by drama, and by narrative computer games. It involves three components (Ryan 2015): Spatial immersion: a sense of place, of connection to the environment. In a computer game, spatial immersion can be a strategic appreciation of the configuration of space and of how this configuration can be used to reach one’s goals. It can also be the pure joy of being able to move through space, of exploring it and discovering ever new regions. Temporal immersion: an experience that can also be called narrative tension (Baroni 2017). It resides in the burning desire to find out what happens next, and it covers the three fundamental narrative experiences of suspense, curiosity, and surprise (Sternberg 1992). In an interactive environment, temporal immersion can also result from the sense that you have only a limited time to perform certain actions, as for instance when you must kill your enemy before he kills you.

V irtual W orlds  •  169

Emotional immersion: the main form of emotional immersion is empathy, which is an emotion directed at others, such as feeling vicariously happy when good things happen to people or characters you like and sad when bad things happen to them. But as tragedy demonstrates, emotional participation in the sad fate of characters is not deep enough to prevent aesthetic pleasure. In computer games, emotions can also be directed at oneself, such as feeling excited, frustrated, dejected, or elated, depending on one’s success.

In real life we find all four kinds of immersion, though the concept of ludic immersion must be extended to cover not only the playing of games but also absorption in useful tasks. In virtual life, as I will call the experience provided by all kinds of representations, different media provide different kinds of immersion, and we can use the four kinds described above as a criterion for comparing the expressive potential of narrative media (see table 3). This assignment of values describes general affordances rather than rigidly distinctive features, and it does not prevent individual texts from challenging the limitations of their medium. Written narratives score high with respect to the three mimetic kinds: they can provide spatial immersion through the description of places; they provide temporal immersion through suspense, curiosity, and surprise; and they can bring readers to tears over the fate of characters. But since they are not interactive, they score zero on ludic immersion. Movies can achieve the same types of immersion as written narratives: high on spatial, temporal, and emotional (even higher in this respect than written narratives: crying is much more frequent when watching films than when reading books), but no ludic immersion. In its standard, plot-driven manifestation, theater is high on emotional and temporal immersion, but it scores lower than movies and written narrative on spatial immersion for two reasons: the phenomenon of the fourth wall separates the spectators from the stage; and the fact that the spatial point of view cannot be changed reduces the setting to a largely static image. In film, by contrast, the camera can move, and in written narrative the narrator can alter the perspective from which scenes are described. Standard video games are by definition strong on ludic, and they score high on both kinds of spatial immersion: sense of place and strategic. On emotional immersion we have a hung jury. As noted above, games inspire self-directed emotions, which depend on the player’s success. Players may also develop strong affective relations to their avatar in games such as Second Life, where they create their avatar, or in games whose purpose is to take care of a virtual creature such as a pet or a baby. But these games do not follow the

170  •  C hapter 9

TABLE 3. Types of immersion in various media TYPE OF MEDIA

SPATIAL

TEMPORAL

EMOTIONAL

LUDIC (INTERACTIVITY)

Written narrative

High

High

High

None

Film

High

High

High

None

Theater

Low

High

High

None

Video games

High

High

Low

High

VR

High

Low

Medium to High

Medium

standard script of defeating enemies in order to fulfill quests, and in the case of Second Life the player’s emotions for the avatar can be regarded as selfdirected. The kind of emotional immersion that is directed at characters is more problematic, at least in standard competitive games, because nonplaying characters are generally seen not as persons but as obstacles to the achievement of the player’s mission or as providers of help. In one case they must be eliminated; in the other they only matter because of the useful objects or information they can give the player. When gamers rescue princesses, they are motivated not by love, as are the heroes of romances, but only by the desire to progress in the game. Much effort has been recently devoted by game designers to involve players in emotion-rich situations, for instance by creating moral dilemmas: should I kill this character who has helped me before or sacrifice these innocent civilians in order to save my life? But emotional involvement conflicts with the pursuit of game goals: if the player chooses the more ethical option, it may put the continuation of the game in jeopardy.

The Distinctive Features of VR What makes a digital application a VR narrative? Is it a 3D graphic presentation? Is it the extension of sensory dimensions beyond hearing and sight? Is it a panoramic representation that goes beyond a natural field of vision? Or does a project qualify as VR only if it requires a helmet or special glasses? We can reject right away 3D as a distinctive feature. There are many films that must be viewed with special glasses for 3D effects, but they are not generally regarded as VR because they are not interactive. The same argument holds for enhanced sensory dimensions. There are now movie theaters equipped with a technology known as 4DX that uses moving seats to enhance regular films with effects such as indoor rain, explosions, smoke, lightening, wind, and so forth. But 4DX films are not interactive, and

V irtual W orlds  •  171

from a narrative point of view they do not offer anything new, since they consist of effects tagged on a film conceived for normal viewing. The experience of 4DX films is more like a passive ride in an amusement park than an active involvement with a computer-generated world. If we combine interactivity with 3D representation requiring a wearable technology, then the best available implementation of VR is video games. There are indeed a number of VR video games on the market, some of which are adaptations of standard games, like Skyrim, and others are native to the medium, such as a rock-climbing game (The Climb) or a game in which the player climbs Mt. Everest (Everest VR). The question is whether adding 3D to games enhances the experience. I can imagine that some walking simulators would gain from an added spatial dimension; if what matters is a sense of “being there” in a beautiful world, then 3D can only enhance the player’s connection to the gameworld. But serious gamers despise these games because of the limited gameplay, calling them games for people who don’t really like games. On the other hand, if you add 3D effects to a classic shooter game where the player has to kill enemies, then the increased realism may be too much for the player, as Jeremy Bailenson suggests (2018, 210). And if a game takes many hours to play, who wants to wear a helmet or VR glasses for so long? Whether games are better in VR depends on the player’s preferences for certain types of games. This leaves us with the feature of panoramic representation. Since one must turn one’s head or one’s body to look around, panoramic representation is inherently interactive. In contrast to three-dimensionality, interactive panoramic representation can be experienced on a regular computer screen. Here I discuss two examples of this feature that have been proposed as VR applications, though one can debate, for reasons to be spelled out later, whether they are really VR. My point in discussing them is to illustrate what 360-degree representation can do for storytelling.

VR as Panoramic Representation A panoramic representation of a storyworld means that a lot of data must be generated that the user will probably not see. In everyday life we do not look around all the time; rather, we focus our attention on certain points of interest (POI, in the jargon; Tricart 2018). In a narrative film the camera does it for us: it shows where the action is, which means that it follows the characters. It would make no sense for the camera to show the ceiling when the hero and heroine are kissing, or, if it does, it is a calculated effect on the part of the director. In a VR system with panoramic representation there is no director

172  •  C hapter 9

who decides what users should see; they can look around all the time, but if the system relies on a strong narrative script, they will focus on the POIs, and most of the data will be wasted. Alternatively, if users decide to take full advantage of the opportunity to look around, they may get the feeling of missing out on what is important to the plot, because as they take time to explore, narrative time keeps moving forward, and something important may be happening where they are not looking. The art of VR narration thus requires a compromise between the instinctive tendency of users to focus on the heart of the narrative action and their desire to exercise their agency by exploring the scene. Or, put differently, the art of VR narrative must find the right balance between temporal immersion, which relies on interest in the evolution of the storyworld, and spatial immersion, which relies on interest in the environment. One way to avoid the conflict of time versus space is to turn the spatial environment itself into the center of interest, so that the user will find something worthy of attention in all directions. A case in point is Clouds over Sidra (Arora and Milk 2015), a six-minute documentary about a camp for Syrian refugees in Jordan which can be watched in 2D on a computer screen, or in 3D by using a Google Cardboard Viewer or a similar device. The purpose of this film is to show the living conditions of the refugees, the boredom, the hopelessness, but also the determination to prepare for a future life of freedom. For the boys, this determination consists of getting physically fit for the return to Syria or of engaging vicariously in freedom fights through computer games; for the girls, more specifically for the heroine, it consists of making the most of their time in the camp by being good students. Gender divisions remain strong: the boys will not let the girls play computer games or exercise in the gym, while the classroom seems to consist exclusively of girls. But there is one small exception to traditional gender roles: the girls are allowed to play soccer, which they could not do in Syria. Through this scene, the film subtly points to a future when women will have more opportunities open to them. The film consists of several distinct scenes, each focused on a specific place: the desert, the classroom, the living quarters of individual families, and so forth. The user can manipulate the camera, scroll left or right, up or down, and explore the entire scene, while the narrator tells her story: (Desert view) We walked for days crossing the desert to Jordan. The day we left my kite got stuck in a tree in our yard. I wonder if it is still there. I want it back. (Family room view)

V irtual W orlds  •  173

My name is Sidra. I am 12 years old. I am in the fifth grade. I am from Syria and there are problems in the cities. I have lived here in the Za’atari camp in Jordan for the last year and a half. I have a big family. Three brothers. One is a baby. He cries a lot. [Baby crying] I asked my father if I cried when I was a baby and he said I did not. I guess I was a stronger baby than my brother. (Camp outdoors view) We like to keep our beds and books clean and without folds. It is like a game to see who has the nicest books with no folds in the pages. I like cloudy days. I feel I am under a cover, like a blanket. (Schoolroom view) Our teachers sometimes pick students who do not raise their hands to answer, so everybody raises their hand even if they do not know the answer. But I do my homework and I know the answer if she calls on me. [Applause]

There is no doubt that the images are documentary, but the status of the narrator is more questionable. Sidra speaks in fluent English, though with a marked accent which is supposed to give authenticity to her testimony—a technique commonly used in voice-over of foreign speakers in television news. The voice, then, is that of an actress who pretends to be Sidra. But where does the text come from: is it an adaptation of what Sidra may have told orally to an interviewer, or was it written by a scriptwriter who imagines what it is like to be Sidra? The perfect coordination between the images and the narration suggests this second solution. But most spectators will regard the narration as the authentic voice of Sidra. By giving an identity to the narrator, even if this identity is not genuine but acted out, the film strengthens its emotional impact because we are more inclined to be affected by the experience of one particu1 lar individual than by the experiences of thousands of anonymous people. Through its focus on the living conditions of Syrian refugees, the film foregrounds the spatial environment at the expense of time and description at the expense of narration. There are suggestions of a past and of a future that involve changes—the crossing of the desert to reach the camp and the 1. The superposition of an acted-out character performing a made-up text with documentary images raises questions concerning the fictionality or factuality of the film. In chapter 2 I suggested that written narratives can be combinations of factual and fictional segments, provided the two kinds are properly framed (cf. Lincoln in the Bardo). Here we have a multimodal text that combines a fictionalized audio with factual images. But many spectators will regard the voice of Sidra as genuine, and will overlook its fictionality, thus taking the whole text as factual.

174  •  C hapter 9

anticipated return to Syria—but between these two events time is suspended, and life is an endless repetition of daily occupations. This lack of eventfulness, which also means a lack of narrative action, means that the various scenes have no outstanding POIs beyond the environment where they take place. The user does not risk missing something important when exploring the setting. My next example contrasts with Clouds over Sidra on two points: it is a fiction, not a documentary, and it relies on parallel events which cannot be seen at the same time; it is therefore necessary to replay the movie to get a more complete understanding of the action. This example, titled Hard World for Small Things, was commissioned by the VR firm WeVR to filmmaker Janicza Bravo. Hard World focuses on a dramatic, literally life-changing event, rather than on a static, ongoing situation. This event is a fictionalized version of the kind of incident that inspired the Black Lives Matter movement: the shooting of an unarmed Black man by White policemen. The film begins by showing two Black men and a Hispanic woman driving through Los Angeles in a convertible, engaging in small talk and obviously enjoying themselves. The dialogue sounds very natural and seems to be at least partly improvised by the actors. By manipulating the controls, we can get an idea of the kind of neighborhood they are driving through. After a while, the people in the car notice two women they know standing on the sidewalk. They park their car in front of a store, and the male passenger gets out and starts talking to the two women while the driver and the woman passenger remain in the car. At this point there are two conversations going on simultaneously. You can hear them both, but when you focus on one conversation, the other becomes background noise and cannot be clearly followed. A new line of action develops when the driver leaves the car to assist an old lady who is crossing the street, and brings her into the car, promising to take her home. Meanwhile, the female passenger has started to talk on her cell phone, adding a third conversation. Depending on where you are looking, you may or may not notice that the male passenger leaves the two women on the sidewalk and enters the store. But then something happens that you will not miss, even if you are looking elsewhere, because it immediately attracts everybody’s attention and interrupts all the ongoing conversations. Two White men walk toward the car, tell the driver extremely rudely to turn down the music on the radio, and ask him how he got the car: did he buy it or was it given to him? These two men are obviously cops, even though they are wearing civilian clothes. So far the action has been presented in one continuous shot. Then a jump cut occurs, and we are inside the store. From that moment on things happen very quickly. The point of view depends on what you have been looking at before, and you don’t have time to change it. If you have been

V irtual W orlds  •  175

looking at the cops, you will still be looking at them once you are inside the store; you’ll see one of them pull a handgun and shoot; you’ll hear the victim fall down and moan as he is dying. But at first sight you don’t really know who he is and why he was shot, though you may guess that he is the man in the car, because this assumption gives closure to the plot. If you have been looking at the people in the car, you will be looking at the victim after the cut. You will see that after he pays for his purchase, he accidentally bumps into one of the cops, and this collision, which is interpreted as aggression, is what motivates the other cop to shoot. If you want to play detective, you can select a point of view and move in slow motion, shot by shot, to reconstitute the event. The possibility of going back in time and of switching perspective provides the user a (near) total apprehension of the what and how of the incident that no lived experience can provide. I write near total, rather than total, because there is something the user cannot do, neither in this film nor in Clouds over Sidra: modify the distance from which the scene is observed. This limitation is due to the fact that both projects rely on camera-produced images that are stitched together. We can move right or left, up or down in the storyworld, but we cannot get closer to things in order to inspect them in greater detail, nor can we pick up and manipulate objects. These limitations confine the user to the role of observer, preventing the active participation that we find in avatar-based computer games. In contrast to panoramic films, computer games rely on dynamic computer-generated graphics rather than on prerecorded data (except for the noninteractive cut scenes). Slater and Sanchez-Vives have argued that only systems based on computer-generated graphics are truly VR. According to their view, the examples I have discussed are nothing more than a new kind of film technique. In the next section, I discuss projects that let users experience the virtual world from a genuine first-person perspective: the perspective of an embodied, active participant.

VR as Experience of Embodiment One of the strengths of VR systems based on computer-generated graphics lies in their ability to manipulate our experience of our own body. The film Avatar is a case in point. The hero is a disabled soldier confined to a wheelchair, but when he enters the planet Pandora as an avatar created by VR he has full control over his virtual body. By manipulating the brain’s image of the body, VR is able to lead to new experiences of embodiment. It can create an out-of-body experience by projecting to users an image of their own body; it can allow

176  •  C hapter 9

them to see themselves from the back; and it can place them in entirely different bodies: for instance into a Black body if they are White, a disabled body if they are abled, and an abled body if they are disabled. One of the first attempts to use VR for an artistic purpose was an electronic art installation named Placeholder, created around 1994 by Brenda Laurel and her team (Ryan 2001). Placeholder was never fully implemented, and even the parts that were realized fell far short of the creator’s vision, but the idea was to make the participant rediscover the world through the body of a mythical creature inspired by Native American myths. If you were Crow, you would have enhanced vision; if you were Spider, you would have eight eyes capturing eight different points of view; if you were Snake, you could see in the dark; and if you were Fish, you could see underwater. One recent project that explores this ability of VR to provide alternative experiences of embodiment is an installation sponsored by the Goethe Institut in Prague based on Franz Kafka’s story The Metamorphosis. It is called VRwandlung, a pun that blends the German title of Kafka’s story (Die Verwandlung) with the VR initials. The project puts the user in the situation of Gregor Samsa, who wakes up one day to discover that he has been transformed into a giant insect. A major theme in Kafka’s story is Gregor’s progressive acquaintance with his new body, how he uses it to move around his room, and how space is reconfigured to fulfill the needs of an insect body. For instance, Gregor can now hide under the bed to spare his mother the awful sight of his body, and he can crawl on the walls to entertain himself. Thanks to VR’s ability to place users into virtual bodies, which they can activate through the movements of their physical body, Kafka’s story is uniquely suited to the affordances of the medium. Users should be able to experience what it feels like to control a body with six legs and giant antennae that impede their movements. They should be able to explore the room. And if they move in front of the mirror, they should discover their new body from a third-person perspective. There is no mirror in Kafka’s story, and Gregor only sees his body from a first-person perspective, but this departure from the text intensifies users’ experience of discovering their new body. One thing the installation cannot do, however, is recreate the plot of the story in its entirety, the way a text or a movie could. If users are going to have some agency, they cannot be forced by the system to behave exactly as Gregor does in Kafka’s story. VRwandlung should not be considered a retelling of Kafka’s story but rather an illustration. Just as the pictures in a storybook propose visual interpretations of some aspects of the story they illustrate, but without capturing the whole plot, so does the VR adaptation of Kafka’s story: it may deepen the experience of people who already know the story, but it cannot replace the text.

V irtual W orlds  •  177

VR as Disembodied Experience I haven’t experienced VRwandlung firsthand (the project is no longer running), and my discussion represents what I imagine VR technology could do with Kafka’s story, rather than what the project actually does. The degree of interactivity that my imagination attributes to VRwandlung may, however, be out of reach for “full VR,” as my next example (one I have actually experienced) suggests. This example is Alejandro Iñárritu’s VR installation Carne y Arena (Virtually Present, Physically Invisible), which represents the experience of migrants trying to cross the US-Mexico border in the Arizona desert while being chased by the border patrol.2 The project was created by digitally capturing the movements of migrants who re-enacted their experience in a studio and by turning this data into virtual characters. It develops in three acts: first, you are led into an empty room with cold, blaring light where you are asked to remove your shoes and leave your possessions. Sandy shoes are strewn on the floor; a sign tells you that they were found on the Arizona-Mexico border, leading to the chilling thought that they could belong to people who died while attempting to cross. After a five-minute wait you are ushered into a space where you experience a six-and-a-half-minute VR simulation of a desert landscape where migrants are pursued by the border patrol. The simulation puts you in a dark desert landscape; gradually, migrants come into view; then a helicopter hovers over them in a deafening noise; the people scamper around, terrorized, while fragments of dialogue are heard, in both Spanish and English. The helicopter moves away, and a period of calm ensues. The migrants are now sitting around a table and eating; a mother is feeding a child. Then the police return, yell orders, point guns, and unleash attack dogs. Chaos breaks out. You can move around the scene and observe the events from various angles and at a variable distance; when you come too close to the walls, an attendant gently pulls you back and redirects you toward the center of the room. After a while you are blinded by an intense light; you try to turn away, to regain your sight; but you are frozen in place. At this point you realize that you are one of the migrants and that you have been caught in the headlights of a border patrol car. The transformation of the user from relatively safe observer to totally helpless participant constitutes the climax of a mimetically enacted narrative. This moment is terrifying, but it does not last long. Light returns to a beautiful desert landscape, and the simulation is over. You exit the installation 2. The trailer, which can be seen at https://www.youtube.com/watch?v=zF-focK30WE, shows how the project was created rather than attempting to give the viewer an idea of the experience.

178  •  C hapter 9

through a long, dark gallery where you can see pictures of the people featured in the simulation and read about their stories. While the mimetic simulation enacts an experience common to many refugees, that of crossing a dangerous space while being pursued by enemies in the sky, the diegetic narratives of the gallery connect this experience to a specific historical situation, that of the Trump administration’s attempt to “defend the border” by building a wall. All these narratives have a happy ending—none of them ends with the hero being captured and deported—which brings relief to the user but also casts doubt about the representative value of the sampling, since very few migrants are able to start a new life in the US. Yet, for all its dramatic force, Carne y Arena does not turn the user into an embodied, active participant. As Adriano d’Aloia (2018) observes, the user “cannot see her hands if she raises them in front of the headset, nor can she see her feet if she tilts her head downwards.” The sense of presence of the characters is so intense that you will walk right through them. By limiting users to a mere point of view, Carne y Arena justifies its subtitle, “Virtually Present, Physically Invisible.” The low throughput of the installation (it allows three visitors every fifteen minutes), extensive data requirements, and psychological demands on the user, all of which restrict the length of the experience to a few minutes, are typical of such projects. These features make it unlikely that full VR will become a major narrative medium allowing complex interpersonal relations and character depiction any time soon, no matter how intense the immersion.

The Immersivity of VR How does VR narrative score with the four types of immersion I have defined? Based on the very limited corpus I have examined, I rate the technology as follows (see the last row of table 3): MEDIUM on ludic immersion: All the projects allow a mild degree of user agency, but this agency does not affect the plot, and it does not compare to the holodeck or to standard computer games. HIGH on spatial immersion: All the projects foreground the experience and exploration of space. LOW on temporal immersion: Neither Sidra nor VRwandlung can sustain a plot that develops in time; and in Hard World it is necessary to replay the movie several times and to alter the perspective to fully understand what is going on. Carne y Arena leads to a climax, but the limited length of full VR projects is not conducive to narrative development.

V irtual W orlds  •  179

MEDIUM to HIGH on emotional immersion: Most advocates of VR (Bailenson 2018; Slater and Sanchez-Vives 2016) stress its potential to create empathy for characters. There is no doubt that if you see a person in 3D, as if this person were present, you will be more deeply touched by her story than if you see a flat image. All the discussed examples create empathy for the characters by giving the participant a sense of what it is like to live under certain conditions, though for Clouds over Sidra and Hard World for Small Things this seems to be due more to the subject matter than to the technology. Researchers have claimed that users feel greater empathy watching an interactive panoramic film like Clouds over Sidra than watching a regular movie (Bosworth and Sarah 2018), but I find the manipulation of the interface distracting, and quite often it provides a view that has little relevance, such as looking at the ceiling or at the floor. It is only in the Carne y Arena example that the technology clearly enhances emotional participation.

Conclusion: The Storytelling Potential of VR In October 2017 Mark Zuckerberg announced the goal of getting “one billion users in virtual reality”—a goal that seemed disproportionate to the sales of the Oculus Rift VR helmet, which had reached only about one million at the time.3 It is certainly not with narrative applications like the ones I have described that Zuckerberg hopes to reach that number, but rather with presence effects that connect users to remote people and venues, such as the sites of rock concerts and sports events. In the age of COVID-19, this is a powerful selling point. VR has also great potential for tourism: it will allow people to visit exotic landscapes and to explore famous monuments without leaving the comfort of their home. And there is the always powerful lure of pornography. But a bad omen for the future of narrative in VR was the announcement, in May 2017, that Facebook was shutting down the Oculus story studio, whose total output was three short films. 3. In response to the question “How many VR headsets have sold in 2020” (this includes all brands, of which Oculus is only a small proportion), Wikipedia answered on August 27, 2020: Estimates suggest that in 2020, sales of augmented and virtual reality (AR/VR) headsets will reach around 5.5 million units. Forecasts project massive growth in both AR and VR headset sales in the coming years, with both technologies combined expected to sell over 26 million units per year by 2023. But the same question in 2021 yielded a projection of only 6.1 million—despite the boost provided by COVID-19.

180  •  C hapter 9

Measured against the technological overhead they require, the narrative examples I have discussed remind us of the mountain that gave birth to a mouse. The storytelling ability of VR should not be taken for granted, because this ability does not necessarily increase with the number of sensory dimensions. What Marshall McLuhan called “cold media”—media poor in sensory dimensions that leave a lot to the imagination, such as written language—can be better storytellers than “hot media,” of which VR is the ultimate. The question of the narrative potential of VR is profiled in a TED talk on YouTube in which Anthony Geffen promotes this “amazing new medium.” The talk ends like this: VR has a huge potential in terms of being a storytelling medium . . . We have the technology, but it’s the story that’s going to drive this medium. Let the storytellers push this medium. This is without doubt a step in human evolution . . . Let the storytellers take up the opportunity in what I think will be a very [inaudible] medium.

Who are those storytellers who are going to drive the medium? It could be that by storytelling Geffen means simply “content providing”—this means any type of content. In journalism, a story is any topic that can stir interest. Alternatively, if we take storytelling in a narrow sense, then a parallel could be drawn between the current state of VR and the early days of film: when film was first invented by the brothers Lumière, it was mainly used to record theater, and it took a while for filmmakers to understand how to narrate in this new medium. As Bosworth and Sarah observe (2018, 2–3) the film Citizen Kane is widely considered by film critics to constitute the turning point when film acquired its full and unique storytelling potential. According to its advocates, VR needs to wait for its Citizen Kane moment: we just need to be patient. But according to another interpretation, the distinctive affordances of VR are not necessarily good at telling stories. This is the opinion of a commentator on Geffen’s talk who goes by the name Cannibal Kid: I’m not sure, if “storytelling” will work properly in this medium. To me, the most astonishing VR-experiments are more “experiences” to me. There is nothing wrong with that, and it works fine. But I kinda want to see what is beneath the craze of “oh my god I’m in a submarine” or: “Dude, I’m standing on Mt. Everest.” Will games with a complex story work? Or will they make gaming just more attractive to the “non-gamers,” because of this new technology? It’s gonna be interesting.

V irtual W orlds  •  181

Cannibal Kid’s comment suggests two different conceptions of narrative. The dominant narratological conception of what makes a media artifact a narrative is the representation of a sequence of events that follow a pattern of conflict and resolution within a storyworld. In this conception narrativity is a matter of plot. More recently, another conception has become popular among narratologists, a cognitive conception based on the notion of experience. According to narratologists such as Monika Fludernik (1996) and David Herman (2009), what makes an artifact a narrative is not the plot but the artifact’s ability to give its audience a sense of “what it is like” to experience a certain situation. Whether or not we accept this conception of narrative, and throw away the notion of plot (something I am not ready to do), it describes very well the examples I have presented. All of them give the user a sense of “what it is like”—what it is like to be a refugee in a camp, what it is like to have a carefree outing with friends interrupted by a senseless shooting, what it is like to have the body of an insect, what it is like to cross the Mexico-US border while being pursued by the border patrol—though in the first two cases, as already noted, it may be due more to the subject matter than to the technology. The strength of VR as a medium of entertainment is not to tell stories with complex characters, like Hamlet, nor stories that live from the plot, like “Little Red Riding Hood,” but to take us to mountaintops and under the sea, to let us fly or walk on the moon, and to give us new bodies. There are good reasons why VR developers prefer to call their projects experiences rather than narratives (Marantz 2016, 88). But I do not wish to say that narrative and VR are incompatible. Even if experience is primary, it can be greatly enhanced if there is some kind of narrative design that motivates the user’s exploration of the virtual world. For instance, a climb of Mt. Everest can be made more exciting if a storm hits and the user has a choice of continuing the ascent or rescuing other climbers. If you are exploring the world under the sea, you may run out of oxygen or have to dodge a shark attack. Or it may be a surprise effect, as when the users of Carne y Arena suddenly realize that they are being caught by the border patrol. The potential of VR as a narrative medium depends on whether creators can put narrative effects in the service of experience, by instilling suspense, dramatic progression, or a sense that the outcome depends on the user’s actions.

CHAPTER 10

Transmedia Worlds Industry Buzzword or New Narrative Experience?

Is there such a thing as transmedia storytelling? In a sense I am playing devil’s advocate. The culture of the past twenty years has produced a vast number of cult narratives that have generated adaptations in many different media, have inspired tens of thousands of texts of fan fiction, and have been continually expanded through action figures, toys, T-shirts, mugs, and other gimmicks. Narrative systems such as George Lucas’s Star Wars, J. R. R. Tolkien’s Lord of the Rings, and George R. R. Martin’s A Song of Ice and Fire keep sprouting all the time. But while we cannot deny the existence of a cultural phenomenon known as transmedia storytelling, we can ask whether it is a form of storytelling or primarily a marketing strategy, whether it is really new, what its various forms are, and what narratology can do about it beyond acknowledging its existence. Henry Jenkins defines this phenomenon as “the flow of content through multiple media platforms” (2006, 2). The dynamics inherent in the metaphor of flow suggests that content evolves as it passes from one medium to another, configured by the particular properties of the various platforms on which it comes to rest before moving on toward another medium. Yet Jenkins also describes transmedia storytelling through a formula that suggests another visualization: “A narrative so large, it cannot be covered in a single medium” (95). Here we imagine various media eagerly reaching toward a fixed content to grab a piece of it. 182

T ransmedia W orlds  •  183

New media promoters tend to regard transmedia storytelling as something radically new and revolutionary, if not as the narrative form of the future, but historians can put this claim in perspective by pointing out the dissemination of Greek myths through various artistic media—sculpture, architecture, drama, epic—or, closer to us, the multiple modes of distribution of Bible stories in the Middle Ages. These stories were not only written in books, which most people could not read, they were also retold orally during sermons, enacted in passion plays, illustrated through painting and stained-glass windows, and even inspired the physically active phenomenon of the Stations of the Cross, where pilgrims relived the passion of Christ by following a fixed itinerary dotted with little chapels. This kind of multimedia treatment is typically reserved for those narratives which are considered foundational for the identity of a group. In the age of globalization, the community-building function of narrative has been taken over by stories like Star Wars, Lord of the Rings, and The Matrix—stories that transcend linguistic, national, and religious boundaries.

The Primacy of Worlds The label transmedia storytelling suggests that narrative content forms a unified story, which means a self-contained type of meaning that follows a temporal arc leading from an initial state to a complication and resolution. This arc is what Aristotle had in mind when he described stories as having a beginning, a middle, and an end (Poetics 5.1; 1996, 13). But story arcs do not lend themselves easily to fragmentation and dispersion into multiple documents. Imagine how annoying it would be to read the beginning of a story in a novel, then to have to go to a movie theater to get the next episode, then to have to buy a comic book, and finally to have to play a computer game in order to find out how it ends. This is not how transmedia storytelling works. Transmedia storytelling is not a serial; it tells not a single story but a variety of autonomous stories, or episodes, contained in various documents and held together by the fact that they take place in the same storyworld. People are willing to look for information in many documents and across multiple platforms because they are so in love with the storyworld that they cannot get enough information about it. In its classical form (i.e., the one represented by commercial franchises), transmedia storytelling is not a game of putting a story together like a jigsaw puzzle; it is instead a return trip to a favorite world. It satisfies the encyclopedist’s passion for acquiring more and more knowledge about a world, or the collector’s passion for acquiring more and more sou-

184  •  C hapter 10

venirs, but not the detective’s passion for reconstructing a story out of disseminated facts. The label transmedia storytelling is therefore a misnomer: the phenomenon should rather be called transmedia world-building (Weiler 2009, Mark J. P. Wolf 2012). We can rank narrative genres according to the relative prominence of world and plot; for instance, tragedy and jokes are plot-dominant while fantasy and science fiction are world-dominant. In a plot-dominated genre, the storyworld is mainly a container for the characters and their actions, and the plot could easily be moved to another type of world. In a world-dominated genre, by contrast, the plot acts as a path through the storyworld that reveals the diversity of its landscapes, the variety of its biological species, and the particularities of its social system. The more richly imagined a storyworld is from the beginning, the more stories can be told about it, and the more discoveries it offers the user. This is why world-dominated narratives present much better material for transmedia storytelling than plot-dominated ones. As a filmmaker told Henry Jenkins, “When I first started, you would pitch a story because without a good story, you didn’t really have film. Later, once sequels started to take off, you pitched a character because a good character could support multiple stories. And now, you pitch a world because a world can support multiple characters and multiple stories across multiple media” (Jenkins 2006, 116).

Relations Between Worlds Transmedia storytelling is the expansion across multiple media of a phenomenon common in literature that Richard Saint-Gelais (2011) has called transfictionality. In Heterocosmica (1998), Lubomír Doležel describes three kinds of relations between fictional worlds that can be regarded as foundational to both literary transfictionality and transmedia storytelling: expansion, modification, and transposition.1 Expansion extends the scope of the original storyworld by adding more existents to it, by turning secondary characters into the heroes of their own story, and by extending the time covered by the original story though prequels, midquels, and sequels. Doležel illustrates this relation 1. Actually, Doležel calls modification displacement. I prefer modification, because displacement suggests the phenomena covered by transposition.

T ransmedia W orlds  •  185

of expansion with a discussion of Jean Rhys’s Wide Sargasso Sea (1966). This novel follows the life of a minor character from Charlotte Brontë’s Jane Eyre, the character of the first wife of Mr Rochester, who only appears in the original text as a madwoman locked up in the attic. In Wide Sargasso Sea she acquires what she is denied in Jane Eyre, namely the opportunity to tell her life story from her own perspective. Modification “constructs essentially different versions of the protoworld, redesigning its structure and reinventing its story” (Doležel 1998, 207). According to Doležel, this relation obtains between Daniel Defoe’s Robinson Crusoe and J. M. Coetzee’s Foe (1988). In Foe, Robinson does not engage in any of the heroic civilizing feats described in the original novel, and he does not keep a journal: his story is the work of a ghostwriter. Most literary examples of modification follow a counterfactual sequence of events by giving a different destiny to the characters. They answer the question “what if?” Transposition. This operation “preserves the design and the main story of the protoworld but locates it in a different temporal or spatial setting” (Doležel 1998, 206). Doležel’s example is The New Sufferings of Young W., by Ulrich Plenzdorf (1973), a novel that transplants the plot of Goethe’s The Sufferings of Young Werther into the German Democratic Republic of the sixties.

To the three relations described by Doležel, I would like to add quotation. Examples of quotation would be a character in one of the Lord of the Rings movies using a light saber borrowed from Star Wars, or an amateur video on YouTube featuring a character wearing a Darth Vader mask in an action set in an American suburb that has nothing to do with the plot of Star Wars (both examples invented). In this case the imported element is not integrated into the storyworld and the effect is one of dissonance and incongruity. The relation of expansion is much more world-preserving than modification and transpositions (and of course quotation) because it does not require any changes in the original story. This difference makes it tempting to conclude that expansions refer to the same world, while modification and transposition refer to related but different worlds. But there is a problem with this view. Imagine that an author writes a novel, and then another author writes a sequel or prequel: can the two texts be said to refer to the same world? The situation is certainly different from the case of authors who write sequels or prequels to their own text, or who use the device of the return of characters. I have no trouble calling the world of the various Star Wars films the same

186  •  C hapter 10

world—they were all conceived by a team headed by George Lucas.2 But now consider the example of the new adventures of Don Quixote published in 1614 by Alonso Fernández de Avellaneda: does it take place in the same world as the Quixote of Cervantes? When Cervantes found out about Avellaneda’s text, he decided to write his own sequel, and that is the second part of Don Quixote. If all three texts are regarded as referring to the same world, this world will contain logical contradictions, since Don Quixote both performs certain actions—those reported by Avellaneda—and does not perform them, since Cervantes wanted to exclude them from the storyworld. Alternatively, the world of Avellaneda’s Quixote can be considered a modification of the original. It follows that the worlds of transfictional texts can relate to the original world in variable ways: overlap in the case of modifications and transpositions (Cervantes’s vs. Avellaneda’s storyworlds); inclusion in the case of expansion by a different author (that would have been the case if Cervantes had regarded Avellaneda as a legitimate sequel); and the same world, just growing bigger, in the case of expansion by the same author (parts I and II of Quixote). Another operation that creates a relation between storyworlds, but without fusing them into one, is the operation of transmedia adaptation. Since different media have different expressive power, it is virtually impossible for two different media to project the same world. Take for instance the difference between novel and film: in a novel, the thoughts of characters can be represented very explicitly and in great detail; in film, the resources for doing so are much more limited: the mental state of characters must usually be inferred by the spectator on the basis of their actions and facial expressions. Another important difference has to do with appearance. In film, as Seymour Chatman (1990) has observed, we know instantly and fully what characters look like, and this enables us to make inferences about their person. In a novel, by contrast, appearance can be left unspecified, and when characters are first introduced, we may know nothing more about them than a name or a referring expression. When characters are described, the description comes piece by piece, and it leaves lots of blanks to be filled in. In addition to the differences that come from the medium itself, adaptations can differ from the original through each of the three transfictional operations. A common operation is to expand the original world by adding 2. This holds for the first six films, but after the acquisition of the copyrights of the franchise by the Walt Disney Company, George Lucas had no part in the creation of installments 7 through 9. Most viewers will, however, regard these films as referring to the same world as 1 through 6 because of the common label (Star Wars), characters, musical themes, visual appearance of the storyworld, and interplanetary travel through advanced technology. Literary critics, influenced by textualism and its adherence to the formula “one text, one world,” are however more likely than fans to regard expansions by a different author as building a different world.

T ransmedia W orlds  •  187

new characters: for instance, the film version of John Fowles’s The French Lieutenant Woman adds a modern film crew to the nineteenth-century characters borrowed from the novel, and it presents the story of these characters as the script of the film whose production is being depicted. An adaptation can also modify the plot by giving it a different ending, as does the Walt Disney film version of The Little Mermaid, which ends with the mermaid getting the prince, or it can transpose a plot to a different world, as does the film Bride and Prejudice, which places the plot of Jane Austen’s Pride and Prejudice in an Indian setting in order to offer a parody of Bollywood films. But there are also adaptations that seek to minimize the differences, for instance the film Jurassic Park compared with the novel by Michael Crichton, who also wrote the script of the film.3 In a transmedia story system, the most common relation between the various documents is expansion: for instance, if there is a game out of a film, the game may invent a new character for the player to control; or it may focus on an aspect of the storyworld that remains undeveloped in other versions. Carlos Scolari mentions three kinds of expansions: interstitial stories, which take place between the installments of a series (one can also place prequels and sequels in this category, since they expand the time period covered by the original document); parallel stories that take place at the same time as the original story, for instance by following the destiny of a secondary character; and peripheral stories “that can be considered more or less distant satellites of the macrostory” (2009, 598). I would regard as peripheral stories the folklore and legends that relate to the objects of the storyworld.4 The relation of modification is much less common than expansion in transmedia projects because it threatens the integrity of the original storyworld, but it is not unheard of: for instance, the game Star Wars: The Force Unleashed transports the player into a counterfactual world where the dark side of Darth Vader and of the Emperor prevails over the Force, while, in the 3. Jenkins dismisses adaptation as a transmedia operation because he views it as a simple retelling of the same story, while transmedia should add new information to the storyworld: “And for many of us, a simple adaptation may be ‘transmedia,’ but it is not ‘transmedia storytelling’ because it is simply representing an existing story rather than expanding and annotating the fictional world” (2009a). Hutcheon and O’Flynn, by contrast, take a very broad view of adaptation, a view that encompasses “recreations, remakes, remediations, revisions, parodies, reinventions, reinterpretations, expansions, and extensions” (O’Flynn in Hutcheon and O’Flynn 2013, 181). 4. Scolari mentions a fourth category: user-generated content platforms like blogs, wikis, etc. But user-generated content does not belong to the same level of analysis as his other categories, since it can belong to any of them. Moreover, blogs and wikis are not necessarily part of the story system; they may consist of user comments, which evaluate the storyworld from an external perspective but do not contribute to its creation.

188  •  C hapter 10

films, it is the other way around. In this case the relationship between the world of the films and the world of the game is partly a matter of plot—in both cases there is a war going on between the Force and the Dark Side—but mostly a matter of names, of objects, of laws of nature, and above all of visual appearance. It is because a certain character is called Darth Vader, because he looks like the hero of the films, and because the main weapon is light sabers that the world of The Force Unleashed is perceived as a version of the Star Wars universe. Modification is also very prominent in the apocryphal texts created by fans; for instance, there is a whole genre, called slash fiction (Gwenllian-Jones 2005), that changes the sexual preferences of famous film or TV characters. The third relation, transposition, is not compatible with the spirit of today’s transmedia franchises. If the plot of Star Wars were transported into a different environment, for instance into a world of medieval fantasy, nobody would recognize it as being a version of Star Wars, because its plot is one of countless stories that tell about a fight between good and evil.5 What makes the Star Wars storyworld distinctive is not the plot but the setting, and if the setting is changed, the whole storyworld loses its identity. As for the relation of quotation, it could be found in parodies, but it presents an even greater challenge than transposition to the spirit of transmedia storytelling, because by calling into question the unity and autonomy of the storyworld it creates an ironic distance from this world that prevents immersion.

Forms of Transmedia Storytelling Brian Clark, a media developer best known for a Facebook post titled “Trans6 media is a lie,” distinguishes between “East Coast” and “West Coast” transmedia storytelling. West Coast stands for what most of us understand under transmedia, namely the commercial mega-franchises of the Hollywood enter-

5. The most distinctive feature of the Star Wars plot is the motif of the father-son rivalry, a rivalry that climaxes in a confrontation leading to the death of one of them. This motif occurs in many narratives around the world, but it is not nearly as common as the fight between good and evil. I doubt that this motif would be in itself sufficient for an audience to recognize a story set in a medieval setting as being a transposition of the Star Wars plot. The connection could only be made if the similarity on the level of plot were reinforced by visual clues or related names. 6. The most famous claim of this post is that there has been no great transmedia hit. This is arguably true if by transmedia hit one must understand top-down, born-transmedia projects rather than franchises exploiting an independently successful novel or film.

T ransmedia W orlds  •  189

tainment industry, such as Star Wars, Lord of the Rings, Harry Potter, and The Matrix; and East Coast stands for anything that is not West Coast. Here are some possible East Coast forms: Transmedia journalism. Inspired by the school known as “slow journalism,” this form of transmedia documents real-world stories through text, video, photos, and oral testimonies. An example is The Sochi Project, by Rob Hornstra and Arnold van Bruggen, a website that contains a variety of documents depicting the Olympic complex as a Potemkin village: that is, as a luxurious facade that hides human suffering. I personally find transmedia a very appropriate mode of presentation for nonfictional projects because news stories naturally come to us through multiple media: newspapers articles, TV spots, audio, books, and movies. This kind of project can be treated like an archive or a database, which means that the user can pick and choose which media objects to consume. Installations that require the simultaneous use of multiple media, both in the sense of media as means of expression and media as delivery systems. An example of this case is Mapping Ararat, a project that commemorates the attempt to create a new Jewish homeland near Niagara Falls by a nineteenth-century Jewish activist named Mordecai Noah. The installation involves three media: augmented reality technology, which allows the participants to see the buildings of the planned community superposed upon their planned real-world location; an audio guide, which gives background information; and a paper map, which guides them through the installation. In this case all three media must be activated for the story to be properly told. Alternate Reality games (ARGs). In these games, players reconstitute a story like a jigsaw puzzle by following a trail of clues that come to them through various delivery systems: mostly websites, but also cellphones, email, posters in the real world, or even live actors. Here media function as channels of information. It is necessary to follow the entire trail of clues to complete the game (McGonigal 2011). While these games have a narrative core, they are played much more for the sake of problemsolving than out of interest in the story to be discovered. Augmented books. An example is Night Film (2013), by Marisha Pessl. The main physical support of the work is a standard book filled with reproductions of visual documents, for instance fake web pages and newspaper articles. This is multimodality, not transmediality. The transmedia dimension comes from an app that can be downloaded to a tablet or smartphone equipped with a camera. By taking pictures of a bird sym-

190  •  C hapter 10

bol on some of the pages, the user can unlock additional content that could not be printed, or that would be too digressive to include in the book: content such as the heroine playing the piano, interviews with characters, or the reading aloud of a fairy tale. For some reason there are no videos in Night Film, but it is easy to imagine that future augmented books will include film clips and animations. Interactive TV. Here I am thinking of projects that link a TV show to information available through other delivery channels. Such a project has been described by Elizabeth Evans (2015). The English TV show The X Factor, a talent show, came together with an app that delivers interactive content during the running of the show. The app provided behind-thescenes information about the contestants, invited viewers to evaluate the performance, and asked them to predict how the contestant would fare in the competition. The transmediality of this form of interactive TV resides in the double screens required to follow the show: the large screen of the TV and the small screen of the smartphone.

West Coast transmedia includes novels generating films, films inspiring novels, TV series connected to ARGs, comics turned into TV series, computer games adapted into films and vice versa. But since the vast majority of narratives that achieve a certain degree of cultural recognition inspire retellings and adaptations, one must wonder whether there is something new and different about the West Coast model of transmedia storytelling besides the fact that we have now many more media than in the age of print. I will use Jenkins’s oft-quoted definition as a starting point for the discussion of this question: Transmedia storytelling represents a process where integral elements of a fiction get dispersed systematically across multiple delivery channels for the purpose of creating a unified and coordinated entertainment experience. Ideally each medium makes its own unique contribution to the unfolding of the story. (2007; italics original)

By insisting on a coordinated and unified entertainment experience, Jenkins’s definition presupposes a deliberate, top-down distribution of content across many media. The examples of East Coast projects that I have given are all topdown, conceived from the very beginning as transmedia, but most if not all of the great West Coast franchises arise bottom-up, by exploiting the success of an already established monomedial narrative, and they grow in a random, uncontrolled manner through a snowball effect. Take the Star Wars franchise: it is now a narrative system with tight top-down management, but it would

T ransmedia W orlds  •  191

never have gotten off the ground if the first trilogy had not been so successful, and if fans had not begun creating all sorts of stories based on the Star Wars universe. One of the reasons we have so few truly top-down transmedia projects is that producers do not want to take the risk of creating the various media objects and then having the project fizzle. Because it depends on media that are very expensive to produce, such as film, TV series, or computer games, the West Coast model of transmedia is affordable for big corporations, but not for creative individuals.

Top-Down Transmedia Projects: Two Examples Transmedia projects that are conceived top-down from the very beginning are the exception rather than the rule. To illustrate the process of top-down creation, I will discuss two transmedia systems in some detail. My first example, The Matrix, is borrowed from Henry Jenkins’s discussion of this case in Convergence Culture (2006). The three feature films, the first released in 1999 and the other two in 2003, were accompanied by short animé films, comics, and computer games, which had been specially commissioned by the authors and producers, Lilly and Lana Wachowski. The most notable feature of the Matrix franchise is that its individual elements depend on each other for a better understanding of the whole. In transmedia storytelling, the various media stand in a relation of competition as much as they stand in a relation of cooperation, since when a bit of story is given to a medium, it must be taken away from another medium to avoid redundancy. If too much is taken away—in other words, if there is no redundancy—the elements of the system will be incomplete and the user will be frustrated, but if nothing is taken away from the individual parts—if they all tell a full story—the user will have no incentive to explore other documents. The design philosophy of the Wachowskis is clearly aimed at cultivating a desire to look beyond the films, especially when it comes to the relationship between the films and the computer games. For instance, in the film The Matrix Reloaded (2003), the good guys succeed in their final mission because a character named Niobe and her team manage to cut down the electrical power that feeds the machines. This event, which is necessary to the logic of the plot, is not shown in the film, but it is one of the tasks that the player must complete in the game. Or to take another example, there is an episode in The Matrix Reloaded where the main character, Neo, encounters a new character named the Kid who wants to join in Neo’s fight against the machines. From their dialogue one can conclude that they already know each other. The story

192  •  C hapter 10

of their first meeting is not shown in the film, but it is the subject matter of one of the animé shorts titled The Kid’s Story. In both examples, the film shows an event that can only be fully explained through a backstory, even though this backstory is told in another document. The overall structure of the Matrix franchise can be compared to pieces of Swiss cheese. The films present an image of the storyworld that is full of plot holes. The function of the other documents is to fill these holes so that the user can form a more complete, more coherent mental representation of the storyworld. Conversely, the film provides information that completes the stories told by the other documents. Each element of the system depends therefore on other documents for the plugging of its holes. This does not mean that the system can be entered through any of its elements. Take the story of the Kid. In the short film, Neo contacts the Kid on his cellphone, the Kid commits himself to Neo’s cause, he is pursued by the Agents, he throws himself from a roof, and rather than getting killed, he lands into another world, which happens to be the (fictionally) real world, where he meets Neo and Trinity.7 To the spectator unfamiliar with The Matrix and The Matrix Reloaded, this would make no sense, since this spectator would not know who Neo or the Agents are, nor that what passes as the real world is in fact a virtual reality projected by the machines. The plot holes in The Kid’s Story would be far too big to provide an autonomous experience, whereas the plot holes in The Matrix Reloaded are relatively minor. Moreover, most spectators of the feature film will be so distracted from the logic of the story by the special effects that they will not notice the inconsequences. The feature films provide a much more efficient entry point—what is called a mother ship in the TV industry (Jenkins 2006, Mittell 2014)—than the other documents. Since the films received the most publicity, I very much doubt that a significant number of users approached the system through the short films, the comics, or the games. My second example of top-down design is represented by Alpha 0.7: Der Feind in Dir (the enemy within), a German TV project produced in 2010 by 8 Südwest Rundfunk. Alpha 0.7 consists of a main document, a TV mini-serial 7. This episode suffers from the same incompleteness of backstory as the encounter between the Kid and Neo in The Matrix Reloaded: how does Neo know the phone number of the Kid if they have never met? 8. Alpha 0.7 was a commercial failure, due in part to lack of publicity and in part to the fact that the episodes were broadcast late in the evening, though they were later made available on the project’s website. This website has now been taken down. I am familiar with Alpha 0.7 because I spent a year at the Johannes Gutenberg University in Mainz as a visiting scholar contributing to a research project on media convergence. As a member of the project, I was invited to a party for the launching of Alpha 0.7 and met the actors and producers.

T ransmedia W orlds  •  193

that ran in six weekly installments of thirty minutes each, of a series of satellite documents available on the internet, and of a series of radio shows presented as a sequel to the mini-serial. Alpha 0.7 tells a story situated in Germany in the year 2017, a near future with respect to the 2010 release date. In this world, a company named Protecta plans to introduce security systems that take the form of brain-scanning technology. When individuals develop the kind of thoughts that could lead to violent crime, their thinking will be changed through the implanting of a chip in their brain, and they will be turned into harmless persons. The planned system will ensure near-total security for German citizens, but it will violate the individual’s right to privacy. The plot of the TV mini-serial unfolds around a young woman named Johanna who is hired by Protecta as a test subject—the seventh such subject, as the title Alpha 0.7 suggests; all the others have mysteriously disappeared or committed suicide. The company implants a chip in Johanna’s brain, unbeknownst to her, in order to control her behavior. Meanwhile, an underground movement called Apollo fights to maintain freedom of thought. In the course of the plot, Johanna escapes from Protecta, makes contact with Apollo and has the chip removed, only to be captured again, have the chip put back in, and so on, in a series of reversals typical of the thriller genre. In the last episode, she is freed from Protecta, but she is wanted by the police for a murder she attempted when she was under the control of the chip. Her adventures as a fugitive are the subject matter of the radio shows. Alpha 0.7 uses television and radio to tell the story, but it also uses a website to provide supplementary documents about the storyworld. For instance, there is a fictional blog maintained by Apollo that discusses the ethical dilemma of imposing security on the population at the cost of privacy. There is a website on which Protecta presents itself to the public as a benefactor of humankind. This positive image is reinforced by a TV spot in which a convicted rapist praises Protecta’s brain-controlling system for turning him into a law-abiding citizen and saving him from a life in jail. There is the blog of Johanna, who describes, day to day, the changes taking place in her mind and her fear that she is suffering from a mental disease. There is also a fictional TV news spot that reports the disappearance of a character named Stefan Hartmann, who was Alpha 0.1, and a web page that contains the links on Hartmann’s computer. The user is invited to visit these links in order to solve the mystery of Hartmann’s disappearance.9 And, finally, there are links 9. An ARG was foreseen to make users investigate Hartmann’s disappearance, but it was not realized for lack of funds.

194  •  C hapter 10

to a number of real-world documents, such as a page on the website of the US Department of Homeland Security that describes a project about “hostile intent detection,” or websites that discuss recent achievements in brain science, which is getting ever closer to reading minds. The nonfictional documents exist independently of the story, but their reference is redirected from the real world to the storyworld. Their function is to suggest that the fictional world of Alpha 0.7 is not as different as one may think from the real world. Through these documents, Alpha 0.7 sends a message typical of dystopic science fiction: “This is a fictional world, but if you don’t do anything, it will become your world.” As I hope to have suggested through this summary, the interest of Alpha 0.7 lies in the storyworld much more than in the plot. The vast majority of spectators will probably limit their exploration of the storyworld to the TV serial, and they will only get to know those features that can be seen from the trail of the plot, but users who consult the satellite documents will gain a broader view of the storyworld and of its relations to the actual world. The satellite documents play the same role as descriptions in a novel: the plot is realized in the TV and radio shows, but the world is fleshed out in the satellites. All in all, the multiple documents that make up the Alpha 0.7 system fulfill a core condition of successful transmedia storytelling that was formulated by Jason Mittell: they must “reward those who consume them without punishing those who do not” (2014, 272).

Choice of World Most transmedia franchises concern storyworlds that are both fictional and fantastic or futuristic. Is this relation necessary or is it coincidental? I think it is neither: there are good reasons for the predilection of transmedia developers for fantastic worlds, but these reasons are not compelling. Although transmedia could very well develop around works with a realistic setting, or concern real events (as The Sochi Project demonstrates), its preference for ontologically remote worlds can be explained in several ways. First, it takes a lot of cognitive effort to imagine a world very different from ours because we can only use our experience of the real world to a limited extent. When a text mentions a horse, we can imagine it on the basis of our knowledge of real horses, but when a text mentions a ghorf, we have only the information provided by the text to visualize this kind of creature, unless there are illustrations. Since they are harder to imagine than realistic worlds, fantastic worlds have much more to gain from a transmedia and multimodal

T ransmedia W orlds  •  195

representation that depicts them through many different kinds of signs that address different senses. Second, once fans have spent the effort to imagine a strange and remote world, they want a return on their cognitive investment. This return consists of many more texts that provide quick and easy access to a world with which they are already familiar, and in which they can be instantly immersed without having to undergo a laborious period of initiation. Third, transmedia franchises tend to develop around blockbuster films, which are themselves often adaptations of bestselling novels. In this age of cultural and economic globalization, Hollywood favors the kind of stories that will captivate audiences all over the world, especially in China. Fantastic worlds, being disconnected from particular geographic, historical, and social circumstances, have a much better chance of fulfilling this requirement than realistic ones. And, finally, the predominance of fantastic worlds in transmedia franchises can be explained by the fact that these franchises target primarily adolescents and young adults. In our early childhood, we are almost exclusively exposed to stories that take place in fantastic worlds, such as fairy tales, myths, and stories about talking animals or animated objects. It is only later in our mental development that we are given realistic stories to read. The literature for young adults is often a mixture of everyday settings and fantastic events: for instance, high school students with whom the reader can identify meeting vampires and zombies who threaten their familiar world. Since children and young adults are big consumers of toys, action figures, comics, videos, and computer games, it is no surprise that the entertainment industry would target the kind of worlds that stirs their imagination.

How to Build Transmedia Storyworlds One question that I ask myself as a narratologist is what it takes for a fictional world and its stories to fascinate the public’s imagination to such a point that they will want to return to it over and over again and experience it in multiple ways. To answer this question, I studied the discourse of the people who are supposed to know best, the people who write guidebooks to transmedia storytelling. I consulted three of them: The Producer’s Guide to Transmedia, by Nuno Bernardo (2011), A Creator’s Guide to Transmedia Storytelling, by Andrea Phillips (2012), and Storytelling across Worlds, by Dowd, Fry, Niedermann, and Steiff (2013). This last book was the most substantial, but I find the title strange because it suggests transposition, the migration of characters and

196  •  C hapter 10

plot to another world. In the vast majority of cases of transmedia, the storyworld remains constant because it functions as the container that keeps the various stories and their media together. None of the books told me how to build a blockbuster story or storyworld, because this is a matter of talent and of luck, and, as the saying goes, poets are born and not made; but reading the guidebooks provided an excellent opportunity to analyze the discourse of the industry and to distinguish it from the kind of scholarly discourse that narratologists should develop. From these three manuals we can extract a rhetoric that is typical of the discourse of the entertainment industry: • Hyperbolic praise. Thanks to transmedia, storytelling will never be the same. “There’s never been a more exciting time to be a storyteller” (Phillips 2012, xi). Transmedia is “amazing,” “groundbreaking stuff,” and a “fascinating idea” (Bernardo 2011, xviii) that will require radically “new ways of thinking” about developments (Dowd et al. 2013, 35). The consumers are not readers, players, or spectators, but invariably fans, a term that suggests passionate and uncritical devotion to storyworlds. • Capitalist attitude. Narrative material and storyworlds are called intellectual property. The creators own this property, despite the tendency of fans and other writers to steal from it, and it is the owners’ duty to protect their property. The purpose of transmedia storytelling is to “monetize” intellectual property by spreading it across as many media as possible, because each medium has its own devotees and increases the size of the audience. The value of a piece of intellectual property is entirely a function of its popular success: in the entertainment industry, there is no such thing as “success of esteem” and no such thing as transmedia driven only by an artistic vision. As Jenkins, Ford, and Green put it (2013, back dust 10 cover), if your content does not spread, it is dead. • Emphasis on “giving the audience what they want.” In a study of the relations between speaker and hearer, Karl Renner (2010) distinguishes a speaker orientation, through which creators express themselves, asking the audience to adapt to their individuality, from a hearer orientation, through which creators adapt to the desires of the audience. While speaker orientation is typical of high art, which is supposed to shake audiences out of their thinking habits, hearer orientation is typical of popular culture. The rhetoric of transmedia falls squarely in the domain 10. I regard Jenkins’s treatment of transmedia storytelling as more promotional than critical, and therefore as closer to industry than to scholarly discourse.

T ransmedia W orlds  •  197

of hearer orientation. The needs of audiences are expressed through eating metaphors: as one reads on the back cover of Storytelling across Worlds (Dowd et al. 2013), the book gives you the tools to meet the “insatiable demands of today’s audience for its favorite creative property.” The role of the transmedia designer is to create and encourage this craving for more content. • Interactivity as a way to save old media. In the digital age, traditional media such as TV, film, and books need to reinvent themselves in order to survive, and the way to do so is to become more interactive and participatory. It could be said that the process of jumping from one platform or medium to another is a form of active involvement, but this means that the audience of transmedia is by definition active since users have to consult many documents. But the authors of advice books have in mind more substantial forms of interactivity. According to Dowd et al., transmedia “assumes that viewer/users are part of a growing participatory culture that does not desire just to watch but to interact, comment, help shape the course of the content and look for (hidden) answers” (2013, 31). This statement contains two practical pieces of advice. The first, “let users help shape the course of events,” sounds good on paper but in practice could lead to disaster: as Bernardo observes, if you give your audience power over the story, they will get rid of the antagonist, solve all major problems, and erase all the drama (2011, 53). The second piece of advice, “have people look for hidden answers,” is much more feasible. The creation of problems to solve will motivate users to get together and exchange information in order to crack the code. The validity of this advice is demonstrated by the popularity of ARGs and by the intense fan activity generated by particularly hermetic narratives such as Mark Danielewski’s House of Leaves (2000) or S, the brainchild of J. J. Abrams and Doug Dorst (2013). But House of Leaves is a monomedial work; encouraging audiences to solve problems is therefore not a distinctive feature of 11 transmedia storytelling.

The Discourse of the Fans The contribution of the discourse of fans to the development of transmedia franchises has been so well documented, especially by Jenkins in Convergence 11. As for S, it consists of a book filled with flat objects such as postcards and handwritten notes. Do these count as a distinct medium, or as modalities?

198  •  C hapter 10

Culture, that there is really nothing I can do here besides restate its importance. Fan discourse takes two forms: a creative one, manifested by fan fiction, remixes, amateur movies, and participation in cosplay events, and a critical one, manifested by online discussion groups and by comments on Amazon. Both forms demonstrate the power of stories and of their worlds to form communities. If today’s culture is participatory, it is as much in the sense of motivating fans to worship together at the altar of a cult narrative as in the sense of collaborative creation. You do not have to write fan fiction to participate in the Star Wars or Harry Potter communities. Active fan participation can be either a bottom-up, spontaneous phenomenon, such as fan fiction and online discussions, or a behavior dictated topdown by the designers of a system, such as asking users to contribute their own materials about a certain topic. An example of system-controlled participation (though not transmedia) is an experiment in collective storytelling conducted by Penguin titled A Thousand Penguins. Participants were asked to expand or modify the beginning of a story posted on a publicly accessible wiki, but the project resulted in pure chaos rather than in the publishable novel Penguin had hoped to produce.12 While it takes the spontaneous discourse of fans to create the kind of cult narratives that make transmedia development profitable, the relations between fan discourse and industry discourse have often been strained. Jenkins recounts the efforts by the Lucas company to encourage but also to control and limit fan production through the creation of a website, Starwars.com, where some fan creations were displayed, for instance the movies that won the annual fan movie competition, judged by George Lucas himself. But in submitting their work, fans gave up any rights to their intellectual property to the Lucas and now Disney companies. The ambiguous attitude of the industry toward fan creations surfaces in this quote by Jim Ward, an executive in the Lucas company: “We love our fans . . . But if in fact someone is using our characters to create a story unto itself, that’s not in the spirit of what we think fandom is about. Fandom is about celebrating the story the way it is” (quoted in Jenkins 2006, 149). The story the way it is (as created by the Lucas or Disney company) is holy scripture, and it is sacrilegious to change any of it because Star Wars is more than a story, it is a universal religion, the common mythology of the globalized world of the twenty-first century. And like any religion based on holy scripture, the franchise finds it of utmost importance to define a corpus of canonical works. 12. A short discussion of A Million Penguins can be found in the Wikipedia entry for collaborative fiction: https://en.wikipedia.org/wiki/Collaborative_fiction.

T ransmedia W orlds  •  199

When the Walt Disney Company bought the rights to the Star Wars brand name from the Lucas company, it revised the canon in order to keep track of the facts of the storyworld, which had proliferated beyond control, and it only kept the six movies produced by Lucas, plus a TV series and an animated film produced in 2008, The Clone Wars. Everything else was expelled from the canon and recategorized as Star Wars Legends. In the parlance of possible worlds theory, the corpus of the Star Wars Legends represents alternative possible worlds, which means descriptions of what could have been (the “what if ”) as opposed to the canonical texts that represent the facts (the “what is”) of the actual world of the franchise. But just as no religion can prevent alternative cults branching out, the delimitation of a canonical corpus cannot prevent fans from exploring the realm of the “what if ” and from sharing their counterfactual productions. Meanwhile, the reboot of the franchise has brought an explosion of new films, comics, novels, and computer games, all of which are part of the new canon.13

Narratological Approaches How can the discourse of narratology distinguish itself from the discourse of the industry? I am not saying that narratology should entirely reject this discourse; the industry has come up with a vocabulary that can be useful to narratology, such as the terms reboot, mother ship, and tie-ins, and with new practices involving old concepts, such as the contrast between canonical and noncanonical elements. Thanks to the discourse of the industry, it is also permissible again to talk about content, a term considered taboo by New Criticism and deconstruction. Narratology should remain skeptical of the hype created by industry discourse and of its claims of radical novelty, but it should also avoid the temptation to declare that there is nothing new under the sun. As a combination of adaptation and transfictionality, transmedia storytelling has obvious roots in the past, and narratology does not need to start from scratch to deal with it; for instance, the principles through which a storyworld can be expanded in the same medium can also operate across media, and we can apply to transmedia most of what Saint-Gelais (2011) describes as transfictional practices. I am thinking here of principles such as extending the timeline, creating prequels and sequels, telling the story of secondary characters, extending the geography of the storyworld, telling the story from a dif-

13. See https://en.wikipedia.org/wiki/Star_Wars_canon.

200  •  C hapter 10

ferent point of view, and leaving some unresolved issues that can be answered in another narrative. These principles are timeless and medium-independent. One difficulty for a narratological approach to transmedia lies in the size of most commercial franchises. As Jenkins observes, transmedia worlds are usually too large for anybody to know them in their entirety (2006, 95). The sheer number of elements means that we need the kind of “big data” approaches advocated by Franco Moretti (2013) rather than the close reading usually favored by literary scholars. Most approaches have been theoretical (such as mine) or enumerative rather than engaged with individual documents. By enumerative I mean approaches that chronicle the number of tie-ins and the development of a franchise over time but that do not go much further than listing the documents. Examples of this “big data” approach are Colin Harvey’s study of the Dr Who franchise (2014) and Rüdiger Heinze’s study of the Alien universe (2015). To restore close reading it will be necessary to focus on the relations between a limited number of documents, such as the relation between the TV series Lost and the ARG devoted to the show, as Jason Mittell (2014) has done, or the relation between the Star Wars film The Force Awakens and the novel of the same name. A comparison of the novel and the movie should raise questions such as the following: Does the novel help us understand some obscure points in the plot of the movie? Does it convey more information than the film, or does it slavishly reproduce the script in words? Can it stand on its own? How does the order in which users view the film and read the novel affect their experience?14 Given the commercial nature of most transmedia franchises, one possible topic of narratological exploration could be how the mother ship document is designed to open opportunities for narrative and transmedia expansion. Take the example of the plot of The Force Awakens (2015), which can be considered the mother ship in the reboot of the Star Wars franchise. It is number 7 of a series of movies, but it takes place some thirty years after episode 6—approximately the same time span that separates the release of episodes 6 and 7. This means that some of the characters from episode 6 are still alive, namely Luke Skywalker, Han Solo, and Princess Leia, and that the same actors 15 could be used if they, too, are still alive. There is consequently some continuity between episodes 6 and 7, and fans will know that they are located in 14. This may, however, be a purely academic question because nearly all readers will have already seen the film, a fact that the author of the novel may (or may not) have taken into consideration. 15. When the actors are dead, they can be digitally simulated if recent footage of them exists. This was the case for the role of Carrie Fisher as Princess Leia in the ninth film, The Rise of Skywalker (2019).

T ransmedia W orlds  •  201

the same world, a sense strengthened by familiar landscapes, technology, and musical themes. The evil empire of episodes 3 through 6 and its main villain Darth Vader have disappeared, but they have been replaced by equally evil antagonists: the First Order, a Nazi-like organization, and Kylo Ren, the son of Han Solo. Just as the real world produced in succession the Nazis, the Khmer Rouge, and the Islamic State, the Star Wars universe will never stop generating evil powers, and only the public’s waning interest will put an end to the saga. The thirty years between episodes 6 and 7 allows a new generation of characters to take over the plot, assuming that the life span of Star Wars characters is roughly the same as that of humans. The new characters come out of nowhere, and this opens endless possibilities for telling their backstory. Another opportunity for expansion is to tell how the Galactic Empire of the past has been replaced with the First Order of the present. Most of the plot of The Force Awakens revolves around an attempt to find Luke Skywalker. A map revealing his location is hidden in a cute robot named BB-8, which made a terrific toy to sell. The map was secreted in BB-8 by a character in the movie, but where the map comes from and who created it remains a mystery: a plot hole that the Disney Company could be happy to fill though more moneymaking products. The film ends when the heroine Rey connects with Luke Skywalker, creating a sense of closure, but this is also an open ending, since we don’t know what she will ask of him or how Luke can help defeat the First Order. These questions are answered in the next film, The Last Jedi (2017). This analysis remains on the level of transfictionality; it does not address the issue of media selection. Ideally, the medium should be dictated by the nature of the content, but with a project as blatantly commercial as Star Wars it is better to invert the question and have the content dictated by the medium. Rather than asking “What is the best medium to tell how young Poe Dameron became a pilot in the service of the Republic?” developers will ask “We need to reach the audience of superhero comics. What part of The Force Awakens would make a good candidate?” The answer will be the backstory of Poe Dameron, because of all the characters in the movie he has the best credentials to 16 be a superhero: male, young, handsome, dashing, and daring. Another question will be “We need to attract gamers; how do we gamify the plot?” In order to do this, users will have to be put in the role of one of the characters, or perhaps become an entirely new character, and they will be given a series of problems to solve. What kind of problems, and how will the game tie in to the movie? Because the storyworlds of most transmedia fran 16. Star Wars: Poe Dameron is a series of thirty-one comic books published by Marvel between April 2016 and September 2018, written by Charles Soule and illustrated by Phil Noto and later Angel Unzueta.

202  •  C hapter 10

chises are fantastic or science-fictional, and because these genres are traditionally rich in action, it should not be too difficult to gamify their stories, since the medium of the video game relies so heavily on fighting, one of the easiest activities to simulate by computer. Through my discussion of The Force Awakens I hope to have suggested that even the most blatantly commercial franchises can reward a narratological approach. I envision the contribution of narratology to transmedia storytelling as comprising the following parts. (1) A transfictional component that describes how stories belonging to the same storyworld are linked together; this component will assess the consistency of the storyworlds. (2) An adaptive component that studies how narrative content travels across media and how the properties of media affect the stories; this component will ask how the various tie-ins take advantage of the affordances of their medium. (3) A mythical component that studies what makes stories and storyworlds into cult narratives, for popular success is the prerequisite to the development of transmedia franchises. This component may ask why it is that the worlds of most franchises are either fantastic or science-fictional; why comic book superheroes are so popular nowadays; whether, when, and how Star Wars has become a religion; and what are the mythical structures of the Star Wars storyworld. The final component, (4), is an audience behavior component, devoted to what people actually do with cult narratives in general and with today’s transmedia systems in particular. Active audiences are nothing new—there are examples in the Renaissance and seventeenth century of readers playing games based on popular narratives or impersonating fictional characters17—but participation has certainly taken new forms thanks to digital technology. This component will ask such questions as “How many different media do standard audiences consult, compared with highly involved ones? How successful is the transmedia extension of the project?” (Consider The Matrix: if the vast majority of fans limit themselves to the three movies, as I suspect they do, is the project truly transmedia?) It could be said that this audience behavior component is more sociological than strictly narratological, but it is the trademark of postclassical narratology that its borders with other disciplines are no longer watertight.

17. For instance, in the seventeenth century, one of the Italian aristocracy’s favorite forms of entertainment was a board game known as the Labyrinth of Ariosto, which was based on Ariosto’s poem Orlando Furioso. A variant of the board game Chutes and Ladders, this game invited players to impersonate Ariosto’s characters and to recite or retell the poem when they landed on certain “narrative” squares. See Ryan (2009).

T ransmedia W orlds  •  203

Conclusion What in the end is transmedia storytelling: the narrative medium of the twenty-first century, or a marketing ploy enticing the user to consume as many products as possible? It may sound strange to regard transmedia as a medium, since it would be a medium of media. But if we define media as a means of expression, and if by using documents belonging to various media it is possible to create experiences that cannot be achieved with a single medium, then transmedia could very well be regarded as a novel means of expression, which means as a medium in its own right. To achieve this honorary status, transmedia will have to find a way to exploit the resources of the media it uses, and it will have to learn how to distribute narrative content among them without frustrating users, that is, without damaging the integrity of the individual components. At its commercial worst, transmedia is the highly profitable practice of giving audiences more of what they want. At its creative best, it could inspire audiences to leave the comfort zone of their favorite medium in order to get a fuller experience of the storyworld, thereby inviting people to reflect on the expressive power of individual media.

BIBLIOGRAPHY

Abbott, Edwin A., and Dionys Burger. (1884) 1994. Flatland, Sphereland. New York: Harper-Collins. Abbott, H. Porter. 2002. The Cambridge Introduction to Narrative. Cambridge: Cambridge University Press. Abrams, J. J., and Doug Dorst. 2013. S. New York: Mulholland Books. Alber, Jan. 2016. Unnatural Narrative: Impossible Worlds in Fiction and Drama. Lincoln: University of Nebraska Press. Alber, Jan, Stefan Iversen, Henrik Skov Nielsen, and Brian Richardson. 2010. “Unnatural Narrative, Unnatural Narratology: Beyond Mimetic Models.” Narrative 18, no. 2: 113–36. ———. 2012. “What Is Unnatural about Unnatural Narratology?: A Response to Monika Fludernik.” Narrative 20, no. 2: 370–82. Alpha 0.7. n.d. Transmedia project produced by Sebastian Rüttner for Südwest Rundfunk, with Arne Lenk, Oliver Stritzel, Victoria Mayer, and Anna Maria Mühe. http://www.Alpha0.7.de/ [No longer available]. Amis, Martin. 1991. Time’s Arrow. New York: Vintage Books. Aristotle. 1996. Poetics. Translated by Malcolm Heath. London: Penguin. Arora, Gabo, and Chris Milk. 2015. Clouds over Sidra. VR project. Accessed September 3, 2021. https://www.youtube.com/watch?v=mUosdCQsMkM. Ashline, William. 1995. “The Problem of Impossible Fictions.” Style 29, no. 2: 215–34. Austin, J. L. 1970. Philosophical Papers. Edited by J. O. Urmson and G. J. Warnock. Oxford: Oxford University Press. Avatar. 2009. Directed by James Cameron, produced by James Cameron and Jon Landay. 20th Century Fox. 205

206  •  B ibliography

Aylett, Ruth, and Sandy Louchard. 2004. “The Emergent Narrative: Theoretical Investigation.” Proceedings of the Narrative and Learning Environments Conference NILE04, Edinburgh, Scotland. Accessed September 3, 2021. https://www.academia.edu/12236192/The_emergent_ narrative_theoretical_investigation. Bailenson, Jeremy. 2018. Experience on Demand: What Virtual Reality Is, How It Works, and What It Can Do. New York: Norton. Bal, Mieke. 1997. Narratology: An Introduction to the Theory of Narrative. Toronto: University of Toronto Press. Banfield, Ann. 1982. Unspeakable Sentences: Narration and Representation in the Language of Fiction. London: Routledge and Kegan Paul. Barnes, Julian. 1985. Flaubert’s Parrot. London: Picador. Baroni, Raphaël. 2017. La Tension narrative: Suspense, curiosité et surprise. Paris: Seuil. ———. 2021. “Of Mice and Men: A Transmedial Perspective on Fictionality.” Narrative 29, no. 1: 91–113. Barthes, Roland. S/Z. (1970) 1974. Paris: Editions du Seuil. English version: S/Z: An Essay. Translated by Richard Miller. New York: Farrar, Straus and Giroux. ———. 1981. Camera Lucida: Reflections on Photography. Translated by Richard Howard. New York: Hill and Wang. Bell, Alice, and Marie-Laure Ryan, eds. 2019. Possible Worlds Theory and Contemporary Narratology. Lincoln: University of Nebraska Press. Bell, John S. 1989. “Six Possible Worlds of Quantum Mechanics.” In Possible Worlds in Humanities, Arts and Sciences. Proceedings of Nobel Symposium 65, edited by Sture Allén, 359–73. Berlin: De Gruyter. Bernaerts, Lars, and Jarmila Mildorf, eds. 2021. Audionarratology: Lessons from Audiodrama. Columbus: The Ohio State University Press. Bernardo, Nuno. 2011. The Producer’s Guide to Transmedia: How to Develop, Fund, Produce and Distribute Compelling Stories across Multiple Platforms. Lisbon: beActive Books. Bogost, Ian. 2017. “Videogames Are Better Without Stories.” The Atlantic, April 25, 2017. Accessed September 3, 2021. https://www.theatlantic.com/technology/archive/2017/04/ video-games-stories/524148/. Bolter, Jay David. 1991. Writing Space: The Computer, Hypertext, and the History of Writing. Hillsdale, NJ: Lawrence Erlbaum. Booth, Wayne C. (1961) 1983. The Rhetoric of Fiction. 2nd ed. Chicago: University of Chicago Press. ———. 1996. “Where Is the Authorial Audience in Biblical Narrative—and in Other ‘Authoritative’ Texts?” Narrative 4, no. 3: 235–53. Bordwell, David. 1985. Narration in the Fiction Film. Madison: University of Wisconsin Press. ———. 2002. “Film Futures.” SubStance 31, no. 1: 88–104. Borges, Jorge Luis. 1964. “Partial Magic in the Quixote.” In Labyrinth: Selected Stories and Other Writings, 193–96. New York: Modern Library. ———. 1999. “The Garden of Forking Paths.” In Collected Fictions, translated by Andrew Hurley, 119–28. London: Penguin Books. Bosworth, Melissa, and Lakshmi Sarah. 2018. Crafting Stories for Virtual Reality. London: Routledge. Boyd, Brian. 2009. On the Origin of Stories: Evolution, Cognition and Fiction. Cambridge, MA: Harvard University Press.

B ibliography  •  207

———. 2017. “Does Austen Need Narrators? Does Anyone?” New Literary History, 48: 285–308. Bravo, Janicza. 2016. Hard World for Small Things. VR project. Accessed September 3, 2021. https://www.youtube.com/watch?v=VPTPQ_CYRsQ. Brooks, Cleanth. 1947. The Well Wrought Urn: Studies in the Structure of Poetry. New York: Harcourt, Brace and World. Brown, Lloyd A. (1949) 1977. The Story of Maps. New York: Dover. Bruce, Colin. 2004. Schrödinger’s Rabbits: The Many-Worlds of Quantum. Washington, DC: Joseph Henry Press. Bruni, Frank. 2019. “The Wall Is a Symbol of Donald Trump’s Neediness.” New York Times, January 8, 2019. Accessed November 24, 2021. https://www.nytimes.com/2019/01/08/opinion/ trump-speech-border-wall.html. Bucher, John. 2017. Storytelling for Virtual Reality: Methods and Principles for Crafting Immersive Narratives. London: Routledge. The Butterfly Effect. 2004. Written and directed by Eric Bress and J. Mackye Kruger. FilmEngine, BenderSpink, Katalyst. Campbell, Joseph. 2008. The Hero with a Thousand Faces. Vol. 17 of The Collected Works of Joseph Campbell. New York: New York Library. Cannibal Kid. n.d. Comment on Anthony Geffen’s TED talk. Accessed September 3, 2021. https:// www.youtube.com/watch?v=TOtTIWBrtJI. Carrère, Emmanuel. (1986) 2005. La Moustache. Paris: Collection Folio. Chatman, Seymour. 1978. Story and Discourse: Narrative Structure in Fiction and Film. Ithaca, NY: Cornell University Press. ———. 1990. Coming to Terms: The Rhetoric of Narrative in Fiction and Film. Ithaca, NY: Cornell University Press. The Climb. 2016. Video game. Crytek. Cohn, Dorrit. 1999. The Distinction of Fiction. Baltimore: Johns Hopkins University Press. Coover, Robert. 1969. “The Babysitter.” In Pricksongs and Descants, 206–39. New York: Plenum Books. Cortázar, Julio. (1968) 1985. “Continuity of Parks.” In Blow-Up and Other Stories, translated by Paul Blackburn, 63–65. New York: Pantheon. Currie, Gregory. 1990. The Nature of Fiction. Cambridge: Cambridge University Press. ———. 1995. Image and Mind. Film: Philosophy and Cognitive Science. Cambridge: Cambridge University Press. ———. 2010. Narratives and Narrators: A Philosophy of Stories. Oxford: Oxford University Press. ———. 2020. Imagining and Knowing: The Shape of Fiction. Oxford: Oxford University Press. D’Aloia, Adriano. 2018. “Virtually Present, Physically Invisible: Virtual Reality Immersion and Emersion in Alejandro González Iñárrritu’s Carne y Arena.” Senses of Cinema, June 2018. Accessed September 3, 2021. http://www.sensesofcinema.com/2018/feature-articles/virtuallypresent-physically-invisible-virtual-reality-immersion-and-emersion-in-alejandro-gonzalezinarritus-carne-y-arena/. Danielewski, Mark Z. 2000. House of Leaves. New York: Pantheon. Dannenberg, Hilary. 2008. Convergence and Divergence: Plotting Time and Space in Narrative Fiction. Lincoln: University of Nebraska Press. Dawson, Paul. 2015. “Ten Theses Against Fictionality.” Narrative 23, no. 1: 74–100. de Bres, Helena. 2021. Artful Truths: The Philosophy of Memoir. Chicago: University of Chicago Press.

208  •  B ibliography

Dear Esther. 2012. Computer game. Written and produced by Dan Pinchbeck. The Chinese Room. Deutsch, David. 1997. The Fabric of Reality: The Science of Parallel Universes—and Its Implications. London: Penguin Books. Deutsch, David, and Michael Lockwood. 1994. “The Quantum Physics of Time Travel.” Scientific American, March 1994, 68–74. Dick, Philip K. 1961. Counter-Clock World. New York: Random House. Diengott, Nilli. 1988. “Narratology and Feminism.” Style 22, no. 1: 42–51. Ditlea, Steve. 1998. “False Starts Aside, Virtual Reality Finds New Roles.” New York Times, March 23, 1998. Accessed November 24, 2021. https://www.nytimes.com/1998/03/23/business/falsestarts-aside-virtual-reality-finds-new-roles.html. Dodson, Zachary Thomas. 2015. Bats of the Republic. New York: Doubleday. Doležel, Lubomír. 1998. Heterocosmica: Fiction and Possible Worlds. Baltimore: Johns Hopkins University Press. Dowd, Tom, Michael Fry, Michael Niedermann, and Josef Steiff. 2013. Storytelling across Worlds: Transmedia for Creatives and Producers. New York: Focal Press. Eco, Umberto. 1979. The Role of the Reader. Bloomington: Indiana University Press. Electronic Literature Collections. Accessed September 3, 2013. http://collection.eliterature.org/. Eliade, Mircea. 1975. Myth and Reality. Translated by Willard R. Trask. New York: Harper and Row. Emerson, Lori. 2014. Reading Writing Interfaces: From the Digital to the Bookbound. Minneapolis: University of Minnesota Press. Ensslin, Astrid, and Alice Bell. 2021. Digital Fiction and the Unnatural: Transmedial Narrative Theory, Method, and Analysis. Columbus: The Ohio State University Press. Evans, Elizabeth. 2015. “Layering Engagement: The Temporal Dynamics of Transmedia Television.” Storyworlds 7, no. 2: 111–28. Everest VR. 2016. Computer game. Sólfar Studios. Falck, Colin. 1994. Myth, Truth and Literature: Towards a True Postmodernism, 2nd ed. Cambridge: Cambridge University Press. Feinberg, Scott. 2021. “‘Awards Chatter’ Podcast: ‘Sopranos’ Creator David Chase Finally Reveals What Happened to Tony (Exclusive).” The Hollywood Reporter, November 2, 2021. https:// www.hollywoodreporter.com/feature/the-sopranos-david-chase-tony-ending-the-manysaints-of-newark-1235040185/. Ferguson, Niall, ed. 1997. Virtual History: What Could Have Been. New York: Fall River Press. Fforde, Jasper. 2002. The Eyre Affair: A Novel. New York: Viking. Fludernik, Monika. 1996. Towards a “Natural” Narratology. London: Routledge. ———. 2008. “Narrative and Drama.” In Theorizing Narrativity, edited by John Pier and José Ángel García Landa, 355–84. Berlin: De Gruyter. ———. 2012. “How Natural Is ‘Unnatural Narratology’; or, What Is Unnatural about Unnatural Narratology?” Narrative 20, no. 3: 357–70. Foer, Jonathan Safran. 2010. “Here We Aren’t, So Quickly.” The New Yorker, June 14 and 21, 2010, 72–73. Foucault, Michel. 1970. The Order of Things: An Archaeology of the Human Sciences. New York: Pantheon. Fowles, John. (1969) 1980. The French Lieutenant’s Woman. Chicago: Signet Books. Frey, James. 2005. A Million Little Pieces. New York: Anchor.

B ibliography  •  209

Garciá, Patricia. 2015. Space and the Postmodern Fantastic in Contemporary Literature. London: Routledge. Geffen, Anthony. 2016. “Storytelling in Virtual Reality.” TED talk. August 15, 2016. Accessed September 3, 2021. https://www.youtube.com/watch?v=TOtTIWBrtJI. Genette, Gérard. 1980. Narrative Discourse: An Essay in Method. Translated by Jane E. Lewin. Ithaca, NY: Cornell University Press. ———. 1982. Figures of Literary Discourse. Translated by Marie-Rose Logan. New York: Columbia University Press. ———. 1997. Fiction and Diction. Translated by Catherine Porter. Ithaca, NY: Cornell University Press. French ed. 1991. ———. 2004. Métalepses. De la figure à la fiction. Paris: Seuil. Gladwell, Malcolm. 1996. “A Matter of Gravity.” The New Yorker, November 11, 1996, 27–28. Goethe Institut, Prague. 2017. VRwandlung. The Metamorphosis VR. Uploaded December 8, 2017. Accessed September 3, 2021. https://vimeo.com/246502816. Goldsmith, Elizabeth C. 1998. “Lafayette’s First Readers: The Quarrel of La Princesse de Clèves.” In Approaches to Teaching Lafayette’s La Princesse de Clèves, edited by Faith C. Beasley and Katherine Ann Jensen, 30–37. New York: Modern Language Association. Goldstein, Rebecca. 2000. Properties of Light: A Novel of Love, Betrayal, and Quantum Physics. Boston: Houghton Mifflin. ———. 2005. Incompleteness: The Proof and Paradox of Kurt Gödel. New York: Norton. Gomel, Elana. 2014. Narrative Space and Time: Representing Impossible Topologies in Literature. London: Blackwell. Goodman, Nelson. 1978. Ways of Worldmaking. Indianapolis: Hackett. Greene, Brian. 1999. The Elegant Universe. New York: Random House. ———. 2004. The Fabric of the Cosmos: Space, Time, and the Texture of Reality. New York: Knopf. Grice, H. Paul. 1971. “Meaning.” In Readings in the Philosophy of Language, edited by Jay F. Rosenberg and Charles Travis, 436–43. Englewood Cliffs, NJ: Prentice Hall. Gwenllian-Jones, Sara. 2005. “Slash Fiction.” In Herman, Jahn, and Ryan 2005, 536–37. Halliwell, Stephen. 2014. “Diegesis-Mimesis.” In Hühn et al. 2014, 1:129–37. Harvey, Colin B. 2014. “A Taxonomy of Transmedia Storytelling.” In Storyworlds across Media: Toward a Media-Conscious Narratology, edited by Marie-Laure Ryan and Jan-Noël Thon, 278–94. Lincoln: University of Nebraska Press. Heinze, Rüdiger. 2015. “‘This Makes No Sense at All’: Heterarchy in Fictional Universes.” Storyworlds 7, no. 2: 75–92. Herman, David, ed. 1999. Narratologies: New Perspectives on Narrative Analysis. Columbus: The Ohio State University Press. ———. 2009. Basic Elements of Narrative. Malden, MA: Wiley-Blackwell. ———. 2013. Storytelling and the Sciences of Mind. Cambridge, MA: MIT Press. Herman, David, Manfred Jahn, and Marie-Laure Ryan, eds. 2005. Routledge Encyclopedia of Narrative Theory. London: Routledge. Herman, David, James Phelan, Peter J. Rabinowitz, Brian Richardson, and Robyn Warhol. 2012. Narrative Theory: Core Concepts and Critical Debates. Columbus: The Ohio State University Press. Hirsch, Marianne. 1997. Family Frames: Photography, Narrative and Postmemory. Cambridge, MA: Harvard University Press.

210  •  B ibliography

Hogan, Patrick Colm. 2003. The Mind and Its Stories: Narrative Universals and Human Emotions. Cambridge: Cambridge University Press. Hornstra, Rob, and Arnold van Bruggen. n.d. The Sochi Project. Accessed September 3, 2021. http://www.thesochiproject.org/en/chapters/the-summer-capital/. Horstmann, Jan. 2018. Theater-Narratologie. Ein Erzähltheoretisches Analyseverfahren für Theaterinszenierungen. Berlin: De Gruyter. Houellebecq, Michel. 2010. La Carte et le territoire. Paris: Flammarion. Hühn, Peter, Jan Christoph Meister, John Pier, and Wolf Schmid, eds. 2014. Handbook of Narratology, 2 vols. Berlin: De Gruyter. Hutcheon, Linda, with Siobhan O’Flynn. 2013. A Theory of Adaptation, 2nd ed. London: Routledge. Iñárritu González, Alejandro. 2007. Carne y Arena (Virtually Present, Physically Invisible). VR installation. Accessed September 3, 2021. Trailer at https://www.youtube.com/watch?v= zF-focK30WE. Jenkins, Henry. 2006. Convergence Culture: Where Old and New Media Collide. New York: New York University Press. ———. “Transmedia Storytelling 101.” 2007. Confessions of an Aca-Fan, March 22, 2007. Accessed September 3, 2021. http://henryjenkins.org/2007/03/transmedia_storytelling_101.html. ———. 2009a. “The Aesthetics of Transmedia: In Response to David Bordwell (Part One).” Confessions of an Aca-Fan, September 10, 2009. Accessed September 3, 2021. http:// henryjenkins.org/2009/09/the_aesthetics_of_transmedia_i.html. ———. 2009b. “The Revenge of the Origami Unicorn: Seven Principles of Transmedia Storytelling.” Confessions of an Aca-Fan, December 12, 2009. Accessed September 3, 2021. http:// henryjenkins.org/2009/12/revenge_of_the_origami_unicorn.html. Jenkins, Henry, Sam Ford, and Joshua Green. 2013. Spreadable Media: Creating Value and Meaning in a Networked Culture. New York: New York University Press. Johansson, Christer. 2012. “Telling and Showing: A Semiotic Perspective.” In Disputable Core Concepts of Narrative Theory, edited by Göran Rossholm and Christer Johansson, 147–82. Bern: Peter Lang Verlag. Johnson, Steven. 2005, Everything Bad Is Good for You: How Today’s Popular Culture Is Actually Making Us Smarter. New York: Riverhead Books. Joyce, James. (1914) 2005. Eveline. In Dubliners, by James Joyce, 25–29. New York: Bantam 2005. Judovitz, Dalia. 1984. “The Aesthetics of Implausibility: La Princesse de Clèves.” MLN 99, no. 5: 1037–56. Kafalenos, Emma. 2006. Narrative Causalities. Columbus: The Ohio State University Press. Kaku, Michio. 2005. Parallel Worlds: A Journey through Creation, Higher Dimensions, and the Future of the Cosmos. New York: Doubleday. Kirkham, Richard. 1992. Theories of Truth: A Critical Introduction. Cambridge, MA: MIT Press. Klauk, Tobias, and Tilmann Köppe, eds. 2014a. Fiktionalität: Ein interdisziplinäres Handbuch. Berlin: De Gruyter. ———. 2014b. “Telling vs. Showing.” In Hühn et al. 2014, 2:846–53. Knausgaard, Karl Ove. 2013–19. My Struggle. 6 vols. New York: Farrar, Straus and Giroux. Knights, Lionel Charles. (1933) 1970. How Many Children Had Lady Macbeth? An Essay in the Theory and Practice of Shakespeare Criticism. Folcroft, PA: Folcroft Press, 1970. Kripke, Saul. 1963. “Semantical Considerations on Modal Logic.” Acta Philosophica Fennica 16: 83–94.

B ibliography  •  211

Kuhn, Thomas S. 1970. The Structure of Scientific Revolutions, 2nd ed. Chicago: University of Chicago Press. Kunz, Diana. 1997. “Camelot Continued: What If John F. Kennedy Had Lived.” In Virtual History: What Could Have Been, edited by Niall Ferguson, 368–91. New York: Fall River Press. Kuroda, S.-Y. 2014. Towards a Poetic Theory of Narration. Essays of S.-Y. Kuroda. Edited by Sylvie Patron. Berlin: De Gruyter. Labov, William. 1972. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press. Lamarque, Peter, and Stein Haugom Olsen. 1994. Truth, Fiction, and Literature: A Philosophical Perspective. Oxford: Oxford University Press. Landow, George P. 1997. Hypertext 2.0: The Convergence of Contemporary Critical Theory and Technology. Baltimore: Johns Hopkins University Press. Lanser, Susan S. 1999. “Sexing Narratology: Toward a Gendered Poetics of Narrative Voice.” In Grenzüberschreitungen: Narratologie im Kontext / Transcending Boundaries: Narratology in Context, edited by Walter Grünzweig and Andreas Stolbach, 167–83. Tübingen: Narr. ———. 2015. “Toward (a Queerer and) More (Feminist) Narratology.” In Narrative Theory Unbound, edited by Robyn Warhol and Susan S. Lanser, 23–42. Columbus: The Ohio State University Press. Laurel, Brenda, Rachel Strickland, and Rob Tow. 1994. “Placeholder: Landscape and Narrative in Virtual Environments.” Computer Graphics 28, no. 2: 118–26. Lavocat, Françoise. 2016. Fait et fiction: Pour une frontière. Paris: Seuil. ———. 2020. Les Personnages rêvent aussi. Paris: Hermann. Le Guin, Ursula K. 1971. The Lathe of Heaven. New York: EOS / Harper Collins. Lewis, David K. 1973. Counterfactuals. Cambridge, MA: Harvard University Press. ———. 1978. “Truth in Fiction.” American Philosophical Quarterly 15: 37–46. ———. 1979. “Possible Worlds.” In The Possible and the Actual: Readings in the Metaphysics of Modality, edited by Michael J. Loux, 182–89. Ithaca, NY: Cornell University Press. ———. 1983. “Postscript to ‘Truth in Fiction.’” In Vol. 1 of Philosophical Papers, 276–80. Oxford: Oxford University Press. ———. 1986a. On the Plurality of Worlds. Oxford: Blackwell. ———. 1986b. “The Paradoxes of Time Travel.” In Vol. 2 of Philosophical Papers, 67–80. Oxford: Oxford University Press. ———. 2004. “How Many Lives Has Schrödinger’s Cat?” Australasian Journal of Philosophy 82, no. 1: 3–22. Lightman, Alan. 1993. Einstein’s Dreams. New York: Pantheon. Lubbock, Percy. 1921. The Craft of Fiction. Dodo Press reprint. Mann, Thomas. (1924) 1996. The Magic Mountain. Translated by John E. Woods. New York: Vintage Books. Mapping Ararat: An Imaginary Jewish Homelands Project. n.d. Accessed September 3, 2021. http:// www.mappingararat.com/. Mateas, Michael, and Phoebe Sengers, eds. 2003. Narrative Intelligence. Amsterdam: John Benjamins. Marantz, Andrew. 2016. “Studio 360.” The New Yorker, April 25, 2016, 86–94. Margolin, Uri. 1990. “Individuals in Narrative Worlds: An Ontological Perspective.” Poetics Today 11, no. 4: 843–71.

212  •  B ibliography

Matravers, Derek. 2014. Fiction and Narrative. Oxford: Oxford University Press. McGonigal, Jane. 2011. Reality Is Broken. London: Penguin Press. McHale, Brian. 1987. Postmodernist Fiction. New York: Methuen. McIntyre, Lee. 2018. Post-Truth. Cambridge, MA: MIT Press. McLuhan, Marshall. 1996. Essential McLuhan. Edited by Eric McLuhan and Frank Zingrone. New York: Basic Books. Mittell, Jason. 2014. “Strategies of Storytelling on Transmedia Television.” In Storyworlds across Media: Toward a Media-Conscious Narratology, edited by Marie-Laure Ryan and Jan-Noël Thon, 253–77. Lincoln: University of Nebraska Press. Moore, Allan, and Kevin O’Neill. 2000. The League of Extraordinary Gentlemen. La Jolla, CA: DC Comics. Moore, John Robert. 1941. “The Geography of Gulliver’s Travels.” The Journal of English and Germanic Philology 40, no. 2: 214–28. Moretti, Franco. 2013. Distant Reading. London: Verso. Morris, Richard. 1984. Time’s Arrows. Scientific Attitudes toward Time. New York: Simon & Schuster. Morson, Gary Saul. 2003. “Narrativeness.” New Literary History 34: 59–73. Murray, Janet H. (1997) 2017. Hamlet on the Holodeck: The Future of Narrative in Cyberspace. Cambridge, MA: MIT Press. ———. 2020. “Virtual/Reality: How to Tell the Difference.” Journal of Visual Culture 19, no. 1: 11–26. Nafziger, George. 1998. Napoleon’s Invasion of Russia. New York: Ballantine Books. Nahin, Paul J. 1999. Time Machines: Time Travel in Physics, Metaphysics, and Science Fiction, 2nd ed. New York: Springer Verlag. Nielsen, Henrik Skov, James Phelan, and Richard Walsh. 2015a. “Ten Theses about Fictionality.” Narrative 23, no. 1: 61–73. ———. 2015b. “Fictionality as Rhetoric: A Response to Paul Dawson.” Narrative 23, no. 1: 101–11. Niven, Larry. 1971. “All the Myriad Ways.” In All the Myriad Ways, by Larry Niven, 1–11. New York: Ballantine Books. Norris, Christopher. 2000. Quantum Theory and the Flight from Realism: Philosophical Responses to Quantum Mechanics. London: Routledge. Nünning, Ansgar, and Roy Sommer. 2008. “Diegetic and Mimetic Narrativity: Some Further Steps towards a Transgeneric Narratology of Drama.” In Theorizing Narrativity, edited by John Pier and José Ángel García Landa, 312–54. Berlin: De Gruyter. Onega, Susana, and José Ángel García Landa. 1996. Narratologies: An Introduction. London: Longman. Patron, Sylvie. 2016. Le Narrateur: Un Problème de Théorie Narrative. Paris: Lambert-Lucas. ———. 2020. “Introduction.” In Optional Narrator Theory: Principles, Perspectives, Proposals, edited by Sylvie Patron, 1–34. Lincoln: University of Nebraska Press. Pavel, Thomas. 1986a. Fictional Worlds. Cambridge, MA: Harvard University Press. ———. 1986b. The Poetics of Plot: The Case of English Renaissance Drama. Minneapolis: University of Minnesota Press. ———. 2003. La Pensée du roman. Paris: Gallimard. Pawlowitsch, Christina. 2020. “Making See: A Structural Analysis of Mathematical and in Particular Game-Theoretical Writing.” Narrative 28, no. 1: 327–54.

B ibliography  •  213

Pedri, Nancy. 2013. “Graphic Memoir: Neither Fact nor Fiction.” In From Comic Strips to Graphic Novels, edited by Daniel Stein and Jan-Noël Thon, 127–53. Berlin: De Gruyter. Pessl, Marisha. 2013. Night Film. New York: Random House. Phelan, James. 1989. Reading People, Reading Plots: Character, Progression, and the Interpretation of Narrative. Chicago: University of Chicago Press. ———. 2005. Living to Tell About It: A Rhetoric and Ethics of Character Narration. Ithaca, NY: Cornell University Press. Phelan, James, and Henrik Skov Nielsen. 2017. “Why There Are No One-to-One Correspondences among Fictionality, Narrative, and Techniques: A Response to Mari Hatavera and Jarmila Mildorf.” Narrative 25, no. 1: 83–91. Phillips, Andrea. 2012. A Creator’s Guide to Transmedia Storytelling: How to Engage and Captivate Audiences across Multiple Platforms. New York: McGraw-Hill. Pier, John. 2014. Metalepsis. In Hühn et al. 2014, 1:326–43. Pimentel, Ken, and Kevin Teixeira. 1993. Virtual Reality: Through the New Looking Glass. New York: Intel / Windcrest McGraw-Hill. Pinget, Robert. (1968) 1978. The Libera Me Domine (Le Libera). Translated by Barbara Wright. New York: Red Dust. Plato. 1968. The Republic of Plato. Translated by Allan Bloom. New York: Basic Books. Pleasantville. 1998. Directed and written by Gary Ross. Larger Than Life Productions. Pratt, Mary Louise. 1977. Toward a Speech Act Theory of Literary Discourse. Bloomington: Indiana University Press. Prince, Gerald. 2003. A Dictionary of Narratology, 2nd ed. Lincoln: University of Nebraska Press. ———. 2020. “‘Narratology Redux’: Review Essay of Brian Richardson, A Poetics of Plot for the Twenty-First Century: Theorizing Unruly Narratives.” Style 54, no. 3: 267–75. Prior, A. N. 2003. Papers on Time and Tense. Edited by Per Hasle, Peter Øhrstrom, Torben Braüner, and Jack Copelan. Oxford: Oxford University Press. Propp, Vladimir. (1928) 1968. Morphology of the Folk Tale. Translated by Laurence Scott. Austin: University of Texas Press. Rabinowitz, Peter J. 2004. “Music, Genre, and Narrative Theory.” In Narrative Across Media: The Languages of Storytelling, edited by Marie-Laure Ryan, 305–28. Lincoln: University of Nebraska Press. ———. 2005. “Showing vs. Telling.” In Herman, Jahn, and Ryan 2005, 530–31. Rajewsky, Irina. 2020. “Theories of Fictionality and Their Real Other.” In Narrative Factuality: A Handbook, edited by Monika Fludernik and Marie-Laure Ryan, 29–50. Berlin: De Gruyter. Renner, Karl N. 2010. “Die Kooperation von ‘Sprecher’ und ‘Hörer.’” In Festschrift für Klaus Kanzog, edited by Michael Schaudig, 20–38. München: Diskurs film. Rescher, Nicholas. 1991. G. W. Leibniz’s Monadology: An Edition for Students. Pittsburgh: University of Pittsburgh Press. Richardson, Brian. 2001. “Voice and Narration in Postmodern Drama.” New Literary History 32, no. 3: 681–94. ———. 2006. Unnatural Voices: Extreme Narration in Modern and Contemporary Fiction. Columbus: The Ohio State University Press. ———. 2016. Unnatural Narrative: Theory, History and Practice. Columbus: The Ohio State University Press. ———. 2019. A Poetics of Plot for the Twenty-First Century: Theorizing Unruly Narratives. Columbus, Ohio: The Ohio State University Press.

214  •  B ibliography

———. 2021. Response to Gerald Prince. Style 55, no. 2: 219–22. Ricoeur, Paul. 1983–85. Temps et récit. 3 vols. Paris: Seuil. Riggs, Stephanie. 2019. The End of Storytelling: The Future of Narrative in the Storyplex. New York: Beat Media Group. Robbe-Grillet, Alain. 1965a. “On Several Obsolete Notions.” In For a New Novel: Essays on Fiction, translated by Richard Howard, 25–54. New York: Grove Press. ———. 1965b. Two Novels by Robbe-Grillet: Jealousy and In the Labyrinth. Translated by Richard Howard. New York: Grove Press. Ronen, Ruth. 1994. Possible Worlds in Literary Theory. Cambridge: Cambridge University Press. Rorty, Richard. 1982. Consequences of Pragmaticism. Minneapolis: University of Minnesota Press. ———. 1989. Contingency, Irony and Solidarity. Cambridge, MA: Cambridge University Press. Roth, Philip. 2005. The Plot Against America. New York: Vintage. Rothman, Joshua. 2018. “As Real as It Goes: Are We Already Living in Virtual Reality?” The New Yorker, April 2, 2018, 30–36. Rubenstein, Mary Jane. 2014. Worlds Without End: The Many Lives of the Multiverse. New York: Columbia University Press. Run Lola Run (Lola rennt). 1998. Directed by Tom Tykver. X-Filme Creative Pool, WDR, Arte. Ryan, Marie-Laure. 1981. “The Pragmatics of Personal and Impersonal Fiction.” Poetics 10: 517–39. ———. 1991. Possible Worlds, Artificial Intelligence and Narrative Theory. Bloomington: Indiana University Press. ———. 1997. “Postmodernism and the Doctrine of Panfictionality.” Narrative 5, no. 2: 165–87. ———. 2001. Narrative as Virtual Reality: Immersion and Interactivity in Literature and Electronic Media. Baltimore: Johns Hopkins University Press. ———. 2007a. “Narrative and the Split Condition of Digital Textuality.” In The Aesthetics of Net Literature: Writing, Reading and Playing in Programmable Media, edited by Peter Gendolla and Jörgen Schäfer, 257–80. Bielefeld: Transcript/Piscataway, NJ: Transaction. ———. 2007b. “Toward a Definition of Narrative.” In The Cambridge Companion to Narrative, edited by David Herman, 22–35. Cambridge: Cambridge University Press. ———. 2009. “From Playfields to Fictional Worlds: A Second Life for Ariosto.” New Literary History 40, no. 1: 159–77. ———. 2010. “Narratology and Cognitive Science: A Problematic Relation.” Style 44, no. 4: 469– 95. http://marilaur.info/cogninarr.pdf. ———. 2011. “Meaning, Intent, and the Implied Author.” Style 54, no. 1: 29–47. http://marilaur. info/impauthor.pdf. ———. 2015. Narrative as Virtual Reality 2: Revisiting Immersion and Interactivity in Literature and Electronic Media. Baltimore: Johns Hopkins University Press. ———. 2016. “Response to Brian Richardson’s Target Essay ‘Unnatural Narrative Theory.’” Style 50, no. 4: 478–83. http://marilaur.info/unnatural.pdf. ———. 2019a. “Fiction: Serious Business or Play-World for the Imagination ?” Style 53, no. 4: 434–39. ———. 2019b. “From Possible Words to Storyworlds: On the Worldness of Narrative Representation.” In Bell and Ryan 2019, 62–87. ———. 2020. “Facts, Fiction and Media.” In Narrative Factuality: A Handbook, edited by Monika Fludernik and Marie-Laure Ryan, 75–94. Berlin: De Gruyter. Ryan, Marie-Laure, Kenneth Foote, and Maoz Azaryahu. 2016. Narrating Space / Spatializing Narrative. Columbus: The Ohio State University Press.

B ibliography  •  215

Saint-Gelais, Richard. 2011. Fictions transfuges: la transfictionnalité et ses enjeux. Paris: Seuil. Saunders, George. 2017. Lincoln in the Bardo. New York: Random House. Saussure, Ferdinand de. 1966. Course in General Linguistics. Translated by Wade Baskin. New York: McGraw Hill. Schaeffer, Jean-Marie. 1999. Pourquoi la fiction? Paris: Seuil. Schaeffer, Jean-Marie, and Ioana Vultur. 2005. “Mimesis.” In Herman, Jahn, and Ryan 2005, 309–10. Scholes, Robert. 1993. “Tlön and Truth: Reflections on Literary Theory and Philosophy.” In Realism and Representation: Essays on the Problem of Realism in Relation to Science, Literature, and Culture, edited by George Levine, 179–81. Madison: University of Wisconsin Press. Schrödinger, Erwin. (1935) 1980. “The Present Situation in Quantum Mechanics.” Translated by John D. Trimmer. Proceedings of the American Philosophical Society 124, no. 5: 323–38. Scolari, Carlos Alberto. 2009. “Transmedia Storytelling: Implicit Consumers, Narrative Worlds, and Branding in Contemporary Media Production.” International Journal of Communication 3: 586–606. Searle, John. 1969. Speech Acts. London: Cambridge University Press. ———. 1975. “The Logical Status of Fictional Discourse.” New Literary History 6: 319–32. Slater, Mel, and Maria V. Sanchez-Vives. 2016. “Enhancing our Lives with Immersive Virtual Reality.” Frontiers in Robotics and AI, December 19, 2016. Accessed September 3, 2021. https://doi.org/10.3389/frobt.2016.00074. Smith, Barbara Herrnstein. 1978. On the Margins of Discourse. Chicago: University of Chicago Press. ———. 1981. “Narrative Versions, Narrative Theories.” In On Narrative, edited by W. J. T. Mitchell. 209–32. Chicago: University of Chicago Press. ———. 1988. Contingencies of Values. Cambridge, MA: Harvard University Press. Sokal, Alan D. (1996) 2000a. “A Physicist Experiments with Cultural Studies.” Reprinted in The Sokal Hoax, edited by the editors of Lingua Franca, 49–53. Lincoln: University of Nebraska Press. ———. (1996) 2000b. “Transgressing the Boundaries: Toward a Transformative Hermeneutics of Quantum Gravity.” Reprinted in The Sokal Hoax, edited by the editors of Lingua Franca, 11–48. Lincoln: University of Nebraska Press. Spiegelman, Art. (1973) 1986. My Father Bleeds History. Vol. 1 of Maus: A Survivor’s Tale. New York: Pantheon. Stanzel, Franz K. (1979) 1982. A Theory of Narrative. Translated by Charlotte Goedsche. Cambridge: Cambridge University Press. Stefanescu, Maria. 2008. “World Construction and Meaning Production in the ‘Impossible Worlds’ of Literature.” Journal of Literary Semantics 37, no. 1: 23–31. Steinmann, Martin Jr. 1981. “Superordinate Genre Conventions.” Poetics 10, no. 2/3: 243–61. Sternberg, Meir. 1992. “Telling in Time II: Chronology, Teleology, Narrativity.” Poetics Today 13: 463–541. Stewart, Susan. 1978. Nonsense: Aspects of Intertextuality in Folklore and Literature. Baltimore: Johns Hopkins University Press. Stoll, Elmer Edgar. 1934. Art and Artifice in Shakespeare: A Study in Dramatic Contrast and Illusion. Cambridge: Cambridge University Press. Strehle, Susan. 1992. Fiction in the Quantum Universe. Chapel Hill: University of North Carolina Press.

216  •  B ibliography

Swift, Jonathan. (1726) 2001. Gulliver’s Travels. London: Penguin. Tarski, Alfred. 1969. “Truth and Proof.” Scientific American, June 1969, 63–77. Tegmark, Max. 2003. “Parallel Universes.” Scientific American, May 2003, 40–51. Thomasson, Amie. 1999. Fiction and Metaphysics. Cambridge: Cambridge University Press. Thon, Jan-Noël. 2016. Transmedial Narration and Contemporary Media Culture. Lincoln: University of Nebraska Press. Tricart, Celine. 2018. Virtual Reality Filmmaking: Techniques and Best Practices for VR Filmmakers. London: Routledge. “Trump Calls Woodward’s Book ‘A Work of Fiction.’” 2018. Washington Post, September 5, 2018. Accessed September 3, 2021. https://www.youtube.com/watch?v=-gaKZtQ7mUk. von Wright, Georg Henrik. 1967. “The Logic of Action: A Sketch.” In The Logic of Decision and Action, edited by Nicholas Rescher, 121–36. Pittsburgh: University of Pittsburgh Press. Walsh, Richard. 1997. “Who Is the Narrator ?” Poetics Today 18, no. 4: 495–513. ———. 2007. The Rhetoric of Fictionality: Narrative Theory and the Idea of Fiction. Columbus: The Ohio State University Press. ———. 2017. “Beyond Fictional Worlds: Narrative and Spatial Cognition.” In Emerging Vectors of Narratology, edited by Per Krogh Hansen, John Pier, Philippe Roussin, and Wolf Schmid, 461–78. Berlin: De Gruyter. ———. 2019a. “Fictionality as Rhetoric: A Distinct Research Paradigm.” Style 53, no. 4: 397–425. ———. 2019b. “Further Reflections on Fictionality: Rejoinders and Responses.” Style 53, no. 4: 508–30. Walton, Kendall. 1990. Mimesis as Make-Believe: On the Foundations of the Representational Arts. Cambridge, MA: Harvard University Press. Warhol, Robyn R. 2005. “Feminist Narratology.” In Herman, Jahn, and Ryan 2005, 161–63. Weiler, Lance. 2009. “Culture Hacker: Lance Weiler Explains Why Film Makers Should Expand Their Films into a ‘Storyworld.’” Filmmaker, Summer 2009. Accessed September 3, 2021. https://www.filmmakermagazine.com/archives/issues/summer2009/culture_hacker.php. Westover, Tara. 2018. Educated: A Memoir. New York: Random House. What Remains of Edith Finch. 2017. Computer game. Directed by Ian Dallas, produced by Alvin Nelson and Michael Fallik. Annapurna Interactive. White, Hayden. 1981. “The Value of Narrativity in the Representation of Reality.” In On Narrative, edited by W. J. T. Mitchell, 1–23. Chicago: University of Chicago Press. Wikipedia. Entry on “Plot Holes.” https://en.wikipedia.org/wiki/Plot_hole. ———. Entry on “Plot Twists.” https://en.wikipedia.org/wiki/Plot_twist. Wittgenstein, Ludwig. 1981. Tractatus Logico-Philosophicus. Edited by D. C. Pears. London: Routledge. Wolf, Mark J. P. 2012. Building Imaginary Worlds: The Theory and History of Subcreation. New York: Routledge. ———, ed. 2017. Revisiting Imaginary Worlds: A Subcreation Studies Anthology. London: Routledge. ———, ed. 2018. The Routledge Companion of Imaginary Worlds. London: Routledge. Wolf, Werner. 2014. “Aesthetic Illusion.” In Hühn et al. 2014, 1:270–87. Wolterstorff, Nicholas. 1980. Works and Worlds of Art. Oxford: Oxford University Press.

B ibliography  •  217

Wyndham, John. (1956) 1959. “Opposite Number.” In The Seeds of Time, 121–39. Harmondsworth: Penguin. ———. (1961) 1965. “Random Quest.” Consider Her Ways and Others, 131–73. Harmondsworth: Penguin. Zipfel, Frank. 2001. Fiktion, Fiktivität, Fiktionalität. Analysen zur Fiktion in der Literatur und zum Fiktionsbegriff in der Literaturwissenschaft. Berlin: Schmidt. Zuckerberg, Mark. 2017. “We Want to Get a Billion Users in Virtual Reality.” The Verge, October 11, 2017. Accessed September 3, 2021. https://www.theverge.com/2017/10/11/16459636/ mark-zuckerberg-oculus-rift-connect. Zunshine, Lisa. 2006. Why We Read Fiction: Theory of Mind and the Novel. Columbus: The Ohio State University Press.

INDEX

Abbot, Edwin: Flatland: A Romance of Many Dimensions, 15

Austin, J. L., 23, 26

Abrams, J. J., and Doug Dorst: S, 117, 197

autofiction, 54

author: implied, 12–13; as narrator, 63

accessibility relations between worlds, 145

Avellaneda, Alonso Fernández de, 186

adaptation, 186

Aylett, Ruth, and Sandy Louchard, 163

Aesop, 54 aesthetic illusion, 122

Bailenson, Jeremy, 179

aesthetics of triviality, 108

Balzac, Honoré de, 70

Alber, Jan, 13, 42, 112, 151, 158, 159

Banfield, Ann, 58, 60

Allen, Woody: The Purple Rose of Cairo, 5

Barnes, Julian: Flaubert’s Parrot, 105

Alpha 0.7: Der Feind in Dir, 192–94; incorporation of real documents, 194

Baroni, Raphaël, 54 Baroque aesthetics, 106

alternate reality game (ARG), 189, 197, 200

Barthes, Roland, 10, 16, 43, 58; S/Z, 75–80; codes, 75–76; readerly vs. writerly, 76

Amadeus: and voice-over narration, 118 anagnorisis, 91

Beaumarchais, Pierre Caron de: The Marriage of Figaro, 104

antimimetic texts, 13

Beckett, Samuel: The Unnameable, 61, 116; Waiting for Godot, 88

Ariosto: Orlando Furioso, 202 Aristotle, 13n7, 21, 79, 82, 90, 91, 97, 98, 107, 163, 183; and mimesis, 111

Bell, Alice, 153n5 Bell, John S., 143

artificial intelligence, 163 Augustine, St., 154

Bernaerts, Lars, 120

Austen, Jane, 41, 43; Pride and Prejudice, 42, 104, 187

Bernejo, José Ferrer: “Dejen Salír,” 152

Bernardo, Nuno, 195–97

219

220  •  I nde x

Bierce, Ambrose: “An Occurrence at Owl Creek Bridge,” 65

approaches, 75–80; as unrealized possibilities, 82; world approaches, 84–89

Bohm, David, 143

Chase, David, 99n5

Bohr, Niels, 126

Chatman, Seymour, 186

Bordwell, David, 5, 121

cheap plot tricks (CPTs) 90, 91–99; and convention of calumniator credited, 96; as deus ex machina, 97–98; as extraordinary coincidence, 92–95, 104; as false news, 95–96; as interrupted action, 98; in Phèdre, 95–96; in La Princesse de Clèves, 92–95; and low probability, 92; prevalence in certain periods, 104; reader acceptance, 103, 108; as tell-tale letter, 94–95; thematic adequacy, 108; ironisation, 105; and genre, 106; preparatory function, 107

Borges, Jorge Luis: “The Book of Sand,” 157–58; “The Garden of Forking Paths,” 130–32, 157; “Partial Magic in the Quixote,” 158; “Pierre Mesnard, Author of the Quixote,” 81n6 Boyd, Brian, 8, 60 Bravo, Janicza, 174 Brecht, Bertolt, 120; The Three Penny Opera, 105 Bride and Prejudice, 187 brilliant plot twists (BPTs), 91–92 Brontë, Charlotte: Jane Eyre, 72 Brown, Dan: Da Vinci Code, 97, 101–102 Brown, Lloyd, 123 Bruce, Colin, 126, 127, 129 Burns, Ken, The Civil War, 67

Citizen Kane, 100–101, 180 Clark, Brian, 188 Clouds over Sidra, 172–74, 175, 179 Coetzee, J. M.: Waiting for the Barbarians, 71 Cohn, Dorrit, 71 Coleridge, Samuel Taylor, 122 Collins, Suzanne: The Hunger Games, 36

Butterfly Effect, The, 134–35

comics and graphic narratives: mimetic and diegetic elements in, 118–19, 119 figure 5

calumniator credited, 79, 96

computer games, 3, 43, 119, 165; and stories, 166; and VR: The Climb, 171; Everest VR, 171; Skyrim, 171; as walking simulators, 166

camera-eye focalization, 70 Campbell, Joseph, 166 Camus, Albert: The Stranger, 60, 72 Capote, Truman: In Cold Blood, 49, 51, 53, 55 Carne y Arena (Virtually Present, Physically Invisible), 177–78, 179, 181 Carrère, Emmanuel: La Moustache, 100, 155–56, 160 Carroll, Lewis: “Jabberwocky,” 146 Cervantes, Miguel de: Don Quixote, 74, 186 characters, 16, 39, 74–89; as abstract artifacts, 81; as allegories, 87; characterhood as scalar property, 87, 88 figure 3; as collections of semes, 77; as counterparts of other characters, 86; crying for, 75, 86; internal vs. external perspective on, 82–85; narrative functions, 85; nonhuman, 86; ontological completeness, 82–83; philosophical approaches, 84–89; and plot resolution, 97; as possible persons, 85, 87; and proper names, 77–78; in readers’ experience, 86; real people in fiction, 47, 85; role in plot, 78–79; textual

Conan Doyle, Arthur: Sherlock Holmes (in The Adventures of Sherlock Holmes), 47, 74 conspiracy theories, 19 contradictions, 147–50; on episode level, 147– 48; on paragraph level, 148; on sentence level, 148–49 Coover, Robert: “The Babysitter,” 148, 160 Cortázar, Julío: “Continuity of Parks,” 150 cosmology, 124–28; and black holes, 125; quantum or many-worlds cosmology, 136–37, 141–43, 144 counterfactuals and counterfactual narratives, 8, 45–46, 133–34 counterpart narratives, 135–41; “The Lathe of Heaven,” 139–40; “Opposite Number,” 137–39; “Random Quest,” 135–37. See also multiverse narratives Crichton, Michael, Jurassic Park, 187 cult narrative, 198

I nde x   •  221

Currie, Gregory, 30n8, 34, 44, 48, 63, 121

Ensslin, Astrid, 153n5 epistolary novel, 70

d’Aloia, Adriano, 178

Escher, M. C., 150

Danielewski, Mark Z., House of Leaves, 153– 54, 160, 197

Euripides: Medea, 97

Dannenberg, Hilary, 92, 104, 133, 135 Davenport, Guy: “The Haile Serlassie Funeral Train,” 151 de Bres, Helena, 52 Dear Esther, 166 deconstruction, 4, 19, 21, 74, 103, 106, 199 deep fake, 43 Derrida, Jacques, 19

Everett III, Hugh, 127 factual narrative, 38, 116, 173n1 factuality, degrees of, 49 fake news, 39 Falck, Colin, 31 fan discourse, 197–99; vs. industry discourse, 198

Deutsch, David, 129

Faulkner, William: The Sound and the Fury, 12, 61

dialogue, 111, 114

Faust (Marlowe’s and Goethe’s), 80

Dick, Philip K.: Counter-Clock Time, 155–56

dithyramb, 111

fiction and fictionality, 8, 16, 34–57; as command to the imagination, 63; and counterfactuals, 45–46; as deliberate framing, 52; vs. factuality, 8; imagining vs. believing, 51; as imitation of nonfiction, 41; integrationist vs. segregationist approaches, 47; and knowledge, 48; and media, 42–44; make-believe (Walton’s) theory, 37, 40, 44, 50–51, 55; naive theory, 35, 39, 55; and narrative techniques, 49; and notion of world, 39; possible worlds theory, 37–38, 40–41, 44, 45–46, 47–49, 56; rhetorical theory, 35, 39–40, 43, 45, 46, 50, 56; scalar or binary, 49–55; speech act (Searle’s) theory, 36–37, 40, 41, 43, 45, 46, 50, 55; and thought experiments, 45–46

documentary film, 43, 173

Fielding, Henry: Joseph Andrews, 116

Dodson, Zachary: Bats of the Republic, 117

Fisher, Carrie, 200

Doležel, Lubomír, 34, 83, 184

Flaubert, Gustave, 8; Madame Bovary, 64

Dickens, Charles, 104; Little Nell (in The Old Curiosity Shop), 75, 86 diegesis: Genette’s conception, 112; in film theory, 112; Plato’s conception, 110–11 diegetic narration: cognitive implications, 115–17; and distance, 115; use of mimetic elements, 116–17; and user’s act of imagination, 120 Diengott, Nilli, 11 digital entertainment: split between popular entertainment and esoteric experimentation, 166 display text, 42

Dowd, Tom, Michael Fry, Michael Niedermann, and Josef Steff, 195–97 Dr Who, 200 Dr. Seuss, 63n3 Eco, Umberto, 34, 146 Einstein, Albert, 20, 128, 143, 154

Fludernik, Monika, 120, 181 Foer, Jonathan Safran: “Here We Aren’t, So Quickly,” 148–49 Force Awakens, The, 200–202 Force Unleashed, The, 188 Foucault, Michel, 19, 58

Electronic Literature Organization, 164

Fowles, John: The French Lieutenant’s Woman, 147, 160, 187

electronic texts, 117

Frey, James: A Million Little Pieces, 52

Eliade, Mircea, 31 Emerson, Lori, 165

García, Patricia, 153

empathy, 169

Gay, John: The Beggar’s Opera, 105, 105n8

222  •  I nde x

Geffen, Anthony, 180 Genette, Gérard, 10, 58, 59, 64, 66, 113, 117, 150 Gödel, Kurt, 20, 154 Goldstein, Rebecca, 143, 147, 155 Goodman, Nelson, 30 Greene, Brian, 124–129, 133

impossible worlds: 17, 145–46; paradoxical nature, 146–47; reader’s strategies for dealing with, 158–60. See also contradictions; impossible space; impossible time; impossible texts; ontological impossibility Iñarritú, Alejandro, 177

Grice, H. P., 62

indexical conception of actuality, 3, 84

Halliwell, Stephen, 110

Inkubus, 153n5

Hard Work for Small Things, 174–75, 179

interactivity, 162, 165; and TV, 190; interactive paradox, 163; as property of media, 167

Hardy, Thomas: The Mayor of Casterbridge, 42 Harvey, Colin, 200 Heinze, Rüdiger, 200 Hemingway, Ernest, 64; “Cat in the Rain,” 70 Hergé: Prisoners of the Sun, 97 Herman, David: 74, 181; and cognitive narratology, 9, 15; and unnatural narratology, 13; on storyworld, 6 Hintikka, Jaakko, 2 Hirsch, Marianne, 54 Hogan, Patrick Colm, 98 holodeck, 163–165 Homer, 4, 112; The Odyssey, 79; The Iliad, 97 Horstmann, Jan, 120 Hospers, John, 156 Houellebecq, Michel: La Carte et le territoire, 54

inferences in storyworld construction, 147

Iversen, Stefan, 13 “Jack and the Beanstalk,” 133 Jenkins, Henry, 182, 184, 187n3, 190, 191, 192, 196n10, 197, 198, 200 Jenkins, Henry, Sam Ford, and Joshua Green, 196 Johnson, Steven, 105 Joyce, James: Ulysses, 71, “Eveline,” 117, 121 Judovitz, Dalia, 106 Kafka, Franz: Metamorphosis, 100, 176; “A Country Doctor,” 160 Kaku, Michio, 126, 127, 133 Kirkham, Richard, 21n2, 21n3 Klauk, Tobias, and Tillmann Köppe, 35, 116 Knausgaard, Karl Ove: My Struggle, 52, 54

Hutcheon, Linda, and Siobhan O’Flynn, 187, 187n3

Knight, Lionel Charles, 82

hypertext fiction, 164, 165

Kuroda, S.-Y., 58; on non-communicative uses of language, 62

illusion breaking, 119 illustrations, 117, 176 imitation: dramatic, 114 immersion, 3, 48, 56, 117, 120, 122, 161, 165; and media, 169–70; types of, 18, 167–70; and VR, 178–79 impossible space, 150–154; and Moebius strips, 153

Kripke, Saul, 2, 123

La Fayette, Madame de: La Princesse de Clèves, 92–95, 103, 104, 106, 108; Zaïde, 106 La Fontaine, Jean de, 54 Labov, William, 67 Lamarque, Peter, and Stein Haugen Olsen, 26, 30n8, 34

impossible texts, 157–58

Landow, George, 164

impossible time, 154–57; erased past, 156–57; reversed-time stories, 155–156

language: non-imitative nature, 23, 113, 115 Lanier, Jaron, 164

impossible vs. unnatural, 146, 151

Lanser, Susan S., 11

I nde x   •  223

Last Jedi, The, 201

metaphor, visual, 54

Laurel, Brenda, 176

Mildorf, Jarmila, 120

Lavocat, Françoise, 35, 82

mimesis: 6; Aristotle’s conception, 111; as dramatic enactment, 112–113

Le Guin, Ursula K.: “The Lathe of Heaven,” 139–40 Leibniz, Gottfried Wilhelm, 2, 123 Lévi-Strauss, Claude, 75 Lewis, C. S.: Chronicles of Narnia, 133 Lewis, David, 2, 3, 29, 34, 35, 41, 44, 84, 123, 128, 131, 158; on fictionality, 16, 28n6; on philosophical implications of quantum cosmology, 142. See also fiction and fictionality: possible worlds theory; indexical conception of actuality; modal realism Lightman, Alan: Einstein’s Dreams, 131 Lingua Franca, 20, 143 linguistics: Saussurian, 22–24, 75 “Little Mermaid, The,” 187 “Little Red Riding Hood,” 101, 153n5, 181 Lubbock, Percy, 116, 121 Lucas, George, and Walt Disney Company: Star Wars, 2, 3, 4, 5, 88, 133, 182, 183, 185, 186n2, 187, 188, 188n5, 189, 190, 191, 198–202 Magritte, René, 150 Mann, Thomas: Doctor Faustus, 58; The Magic Mountain, 64 Marantz, Andrew, 181 Margolin, Uri, 83 Martin, George R. R.: A Song of Ice and Fire, 182

mimetic narration: cognitive implications, 115–17; limitations, 118; use of diegetic elements, 118–20, 122; illusion breaking in, 119; and sense of unmediated presence, 115; and stage directions, 120; and user’s act of imagination, 121 mimetic texts, 13, 51. See also antimimetic texts minimal departure, principle of, 48, 158; vs. maximal departure, 158 missing fourth wall theory of drama, 121 Mittell, Jason, 192, 194, 200 modal logic, 145 modal realism, 85, 128 Molière: Tartuffe, 97 Moore, Alan, and Kevin O’Neill: The League of Extraordinary Gentlemen, 6, 6n3, 150 Moretti, Franco, 200 Morris, Richard, 155n6, 155n7 multimodality, 117 multiverse narratives, 129–143; “All the Myriad Ways,” 141; The Butterfly Effect, 134– 35; Einstein’s Dreams, 131–31; forking-path narratives, 130–131; wormhole narratives, 133. See also counterpart narratives Murdoch, Iris, 36 Murray, Janet, 163, 165 myth and mythical truth, 31–32; vs. science, 32

Mateas, Michael, and Phoebe Sengers, 163

Nabokov, Vladimir: Pale Fire, 41

Matravers, Derek, 51

Nafziger, George: Napoleon’s Invasion of Russia, 55

Matrix, The, 183, 189, 191–92, 192n7, 201 Maupassant, Guy de, 116 McGonigal, Jane, 189 McHale, Brian, 148, 149 McIntire, Lee, 19 McLuhan, Marshall, 180 media, 5; ability to make truth claims, 42; visual, 54; and interactivity, 167–70, 203. See also transmedia storytelling metalepsis, 17, 150

Nahin, Paul, 134, 154, 156 narratee, 72, 116 narration: figural, 117; present tense, 115; voice-over, 118 narrative turn, 4 narrative: and discourse, 5; as cognitive construct, 5; conversational, 49, 67; defined, 7; and emotions, 103; as experience, 178, 181; forking path, 5, 130; and gameplay, 165; generation in real time, 163; mul-

224  •  I nde x

timodal, 117; natural vs. unnatural, 14, 121; as (non)communication, 12, 116, 121; realistic nature of, 9; and representation, 113; and types of signs, 113; virtual embedded, 9. See also narrativity narrativity: as obstacle to pluralism, 80; as scalar phenomenon, 14 narratology: classical, 73; cognitive, 15; feminist, 10–12; postclassical, 10; rhetorical, 11–12; unnatural, 9, 13–14, 42; and what is, what if, as if approach, 9–15; and transmedia storytelling, 199–203 narrator: 16, 58–73; as abstract textual function, 60; backgrounded, 117; creative, transmissive and testimonial functions, 67–72; individuated, 66; intrusive, 116; as a matter of degree, 59; mandatory vs. optional theories, 59–66; as mediator, 115; morphing, 64; omniscient, 14, 14n9, 59, 60, 121; as oral storyteller, 67; and reader’s act of imagination, 64; relation to author, 65–66; unreliable, 12, 29, 65n7; “thin,” 65; in visual narratives, 61, 72 New Criticism, 4, 74, 75, 103, 199 Newton, Sir Isaac, 152 Nielsen, Henrik Skov, 13, 16, 35. See also fiction and fictionality: rhetorical theory Niven, Larry: “All the Myriad Ways,” 141 novel: New Novel, 78; modernist, 116 Nünning, Ansgar, 120 Occam’s razor, 40, 66, 80, 128 ontological impossibility, 149–50, 154 opera, 119 parallel worlds, 123–44, 151 Path, The, 153n5 Patron, Sylvie, 11, 58–59, 62, 65–66 Pavel, Thomas, 31, 32, 83, 133, 158 Pawlowitsch, Christina, 118 Pedri, Nancy, 54 Peirce, Charles Sanders, 113 Perrault, Charles, 101 Pessl, Marisha: Night Film, 117, 189 Phelan, James, 9, 12, 16, 34, 67, 85, 116. See also narratology: rhetorical; fiction and fictionality: rhetorical theory

Phillips, Andrea, 195 photography, 43 physics: approaches to quantum phenomena, 143; classical vs. modern, 146, 152; Copenhagen interpretation, 126; Einstein’s theory of relativity, 20, 128, 143, 154; Heisenberg’s principle of uncertainty, 20, 143; many-worlds interpretation of quantum phenomena, 17, 123, 126; Schrödinger’s cat, 126–27; string theory, 151; and time, 155; wave function, 126, 129 Pier, John, 150 Pinget, Robert: Le Libera, 148, 160 Placeholder, 176 Plantinga, Alvin, 2 Plato, 4, 13n7, 16, 17; and mimesis/diegesis contrast, 110–114 Pleasantville, 150, 152 plot holes (PHs), 99–102; in Citizen Kane, 100–101; in film, 100; in “Little Red Riding Hood,” 101; in Da Vinci Code, 102 plot, 16, 90–109; artificiality, 104; and circulation of information, 94; as defining element of narrative, 181; use of magic, 98. See also cheap plot tricks; plot holes; brilliant plot twists (BPTs) plural ontology, 132–35. See also cosmology: quantum cosmology possible worlds theory, 2, 16, 28, 51, 84, 123; and many-worlds interpretation of quantum phenomena, 128; narratological applications, 128. See also accessibility relations; fiction and fictionality: possible worlds theory; modal logic postmodern theory, 74 postmodernism, 19 post-truth era, 19 Pratt, Mary Louise, 41–42, 67, 72 Prince, Gerald, 14n8, 59 Prior, A. N., 131 Prokofief, Sergey: Peter and the Wolf, 67 Propp, Vladimir, 109, 166 Proust, Marcel, 4 Proust, Marcel, 4; In Search of Lost Time, 64 Rabinowitz, Peter, 9, 12, 67, 116, 119 Racine, Jean: Phèdre, 95–96

I nde x   •  225

radio drama, 120

signs, types of, 113

Rajewsky, Irina, 35

Slater, Mel, and Maria Sanchez-Vives, 165, 179

reality: as mathematical structure, 125, 152

small world effect, 92, 104

recentering, 39, 44, 85

Smith, Barbara Herrnstein, 7, 24–25, 41, 72

reference: to fictional characters, 81; and storyworlds, 146

Sokal, Alan, and Sokal hoax, 20–22

relativism, 21; and linguistics, 22–24

Sophocles: Oedipus Rex, 91, 107, 108

Renner, Karl, 196 representation: diegetic vs. mimetic, 113–15, 114 figure 4; and imitation, 112–123; 3D, 171; panoramic, 171–75

Sommer, Roy, 120 Sopranos, The, 98–99

Rescher, Nicholas, 2, 123

space: curved topology, 152–53; impossible vs. unnatural, 152; and infinity, 124; nonEuclidean, 151–52; number of dimensions, 124; recursive embedding, 125

Richardson, Brian, 10, 13, 14, 42, 74, 87, 112

speech act theory, 26

Ricoeur, Paul, 8n4, 13

Spiegelman, Art: Maus, 53–54, 55, 118, 119 figure 5

Robbe-Grillet, Alain: La Jalousie, 78; Dans le labyrinthe, 147, 148, 160 Ronen, Ruth, 83

stage directions, 120 Stanzel, Franz K., 117

Rorty, Richard, 22, 25–26, 33

Star Wars: Poe Dameron, 201n16

Roth, Philip: The Plot Against America, 46, 134

Stefanescu, Maria, 146 Steinmann, Martin, 79, 96

Rousseau, Jean-Jacques: Les Confessions, 55

Sternberg, Meir, 76, 164

Rowling, J. K.: Harry Potter, 98, 198

Sterne, Laurence: Tristram Shandy, 65, 116

Run, Lola, Run, 5

Stewart, Susan, 147 Stoll, Elmer, 96

Saint-Gelais, Richard, 7, 184, 199

story: vs. discourse, 5; vs. world, 5

Sartre, Jean-Paul: La Nausée, 72

storyology, 109

Saunders, George: Lincoln in the Bardo, 48, 53, 173

storyworld, 29; against, 7; vs. real world, 38. See also world

Saussure, Ferdinand de, 22, 62, 75; and arbitrariness of signs, 22–23; langue vs. parole, 23

Strehle, Susan, 152

Schaeffer, Jean-Marie, 6, 35, 44; and Iona Vultur, 8, 8 n4, 13, 101n6 Scholes, Robert, 23 Schrödinger, Erwin, 126 science fiction, 144 Scolari, Carlos, 187, 187n3 Searle, John, 12, 16, 34, 35, 36, 41, 43, 45. See also fiction and fictionality: speech act theory

structuralism, 75 Swift, Jonathan: Gulliver’s Travels, 133, 151 Swiss cheese worlds, 160, 192 Tegmark, Max, 124–29, 143, 152 tellability, 92, 107 textual world, see storyworld textualism, 4, 74 theoretical fiction, 129 theory of mind, 104

self-reflexivity, 161

Thomasson, Amie, 34, 81–84

Shakespeare, William: Othello, 79

thought experiments, 45–46

Shaw, George Bernard, 120

Thousand and One Nights, The, 72, 158

showing vs. telling, 113, 116–117; and media, 120

time: and spatial metaphors, 131. See also impossible time

226  •  I nde x

time-travel narratives, 134 Tolkien, J. R. R., 4; Lord of the Rings, 50, 82, 182, 183, 185, 189

and embodiment, 175–76; as experience, 180–81; immersivity, 178–79; narrative potential, 179–81; rebirth, 167

Tolstoy, Leo: War and Peace, 46, 47, 48n13, 50, 55, 85

VRwandlung, 176

transfictionality, 7, 56; as world expansion, 184; as plot modifications, 185, 188; as transposition, 185; as quotation, 185

Walsh, Richard, 7, 16, 35, 49, 86. See also fiction and fictionality: rhetorical theory

transmedia storytelling, 18, 182–203; as alternate reality game (ARG), 189; as augmented book, 189; and concept of world, 183; East Coast vs. West Coast, 188–89; favorite kinds of worlds, 194–95; and “how to” books, 195–97; as installation, 189; as journalism, 189; and media selection, 201; and narratology, 199–203; old or new concept ?, 183; and open plot, 200; and story arcs, 183; top-down vs. bottom-up development, 191–94. See also transfictionality; adaptation; fan discourse

Walton, Kendall, 16, 34, 35, 45, 56, 158; on fictional truth, 29; and representation, 112. See also fiction and fictionality: makebelieve theory Ward, Jim, 198 Warhol, Robin, 9, 11, 74 Weiler, Lance, 183 Wells, H. G.: War of the Worlds, 97 Westover, Tara: Educated, 54 what is, what if, as if approach, 7–9; and dominant schools of narratology, 9–15 What Remains of Edith Finch, 166

Tricart, Céline, 171

White, Hayden, 103–04

truth, 16, 19–33; as coherence, 21; and competition, 31; as consensus, 21, 25–26; as correspondence, 21, 28–29, 33; and discourse genres, 27; and factual narratives, 38; in fiction, 46–49; fictional, 29; as language game, 25, 33; of myth, 31–32; literary, 30–31; postmodern critiques of truth as correspondence, 19–20, 24–26, 28, 33; and religion, 32; in science, 28–29; of statements about fiction, 8; truth claim ability of media, 42; of value judgments, 24

Wiesel, Elie: Night, 55 Wolf, Mark J. P., 4, 184 Wolsterstorff, Nicholas, 82 Woolf, Virginia: Mrs. Dalloway, 12

verisimilitude, 94, 106, 107

world, concept of, 1; as container, 7, 41; cosmological interpretation, 1–2; and computer games, 3; duplicate, 124; as environment, 3; in literary and narrative theory, 4; as mental construct, 41; as network of relations, 7, 41; plurality of, 17, chapter 7; and possible worlds theory, 2; as secondary creation, 4; as transmedia construct, 18; wormhole universe, 151. See also impossible worlds; parallel worlds; storyworld; Swiss cheese worlds; transfictionality; transmedia storytelling

video games, see computer games

Wyeth, Andrew, 3

unspeakable sentences, 60

voice-over narration, 118, 122 Voltaire: Candide, 108 von Wright, Georg Henrik, 8 VR (virtual reality technology), 17, 162–81; decline, 164; defined, 162; distinctive features, 170; and disembodiment, 177–78;

Wyndham, John: “Random Quest,” 134, 135– 37; “Opposite Number,” 137–39 Zipfel, Frank, 35 Zuckerberg, Mark, 179 Zunshine, Lisa, 104

T H E O R Y A N D I N T E R P R E TAT I O N O F N A R R AT I V E JAMES PHELAN, K ATRA BYRAM, AND FAYE HALPERN, SERIES EDITORS ROBYN WARHOL AND PE TER RABINOWITZ, FOUNDING EDITORS EMERITI

Because the series editors believe that the most significant work in narrative studies today contributes both to our knowledge of specific narratives and to our understanding of narrative in general, studies in the series typically offer interpretations of individual narratives and address significant theoretical issues underlying those interpretations. The series does not privilege one critical perspective but is open to work from any strong theoretical position. A New Anatomy of Storyworlds: What Is, What If, As If by Marie-Laure Ryan Narrative in the Anthropocene by Erin James Experiencing Visual Storyworlds: Focalization in Comics by Silke Horstkotte and Nancy Pedri With Bodies: Narrative Theory and Embodied Cognition by Marco Caracciolo and Karin Kukkonen Digital Fiction and the Unnatural: Transmedial Narrative Theory, Method, and Analysis by Astrid Ensslin and Alice Bell Narrative Bonds: Multiple Narrators in the Victorian Novel by Alexandra Valint Contemporary French and Francophone Narratology edited by John Pier We-Narratives: Collective Storytelling in Contemporary Fiction by Natalya Bekhta Debating Rhetorical Narratology: On the Synthetic, Mimetic, and Thematic Aspects of Narrative by Matthew Clark and James Phelan Environment and Narrative: New Directions in Econarratology edited by Erin James and Eric Morel Unnatural Narratology: Extensions, Revisions, and Challenges edited by Jan Alber and Brian Richardson A Poetics of Plot for the Twenty-First Century: Theorizing Unruly Narratives by Brian Richardson Playing at Narratology: Digital Media as Narrative Theory by Daniel Punday Making Conversation in Modernist Fiction by Elizabeth Alsop Narratology and Ideology: Negotiating Context, Form, and Theory in Postcolonial Narratives edited by Divya Dwivedi, Henrik Skov Nielsen, and Richard Walsh Novelization: From Film to Novel by Jan Baetens Reading Conrad by J. Hillis Miller, Edited by John G. Peters and Jakob Lothe Narrative, Race, and Ethnicity in the United States edited by James J. Donahue, Jennifer Ann Ho, and Shaun Morgan Somebody Telling Somebody Else: A Rhetorical Poetics of Narrative by James Phelan Media of Serial Narrative edited by Frank Kelleter Suture and Narrative: Deep Intersubjectivity in Fiction and Film by George Butte The Writer in the Well: On Misreading and Rewriting Literature by Gary Weissman Narrating Space / Spatializing Narrative: Where Narrative Theory and Geography Meet by Marie-Laure Ryan, Kenneth Foote, and Maoz Azaryahu Narrative Sequence in Contemporary Narratology edited by Raphaël Baroni and Françoise Revaz

The Submerged Plot and the Mother’s Pleasure from Jane Austen to Arundhati Roy by Kelly A. Marsh Narrative Theory Unbound: Queer and Feminist Interventions edited by Robyn Warhol and Susan S. Lanser Unnatural Narrative: Theory, History, and Practice by Brian Richardson Ethics and the Dynamic Observer Narrator: Reckoning with Past and Present in German Literature by Katra A. Byram Narrative Paths: African Travel in Modern Fiction and Nonfiction by Kai Mikkonen The Reader as Peeping Tom: Nonreciprocal Gazing in Narrative Fiction and Film by Jeremy Hawthorn Thomas Hardy’s Brains: Psychology, Neurology, and Hardy’s Imagination by Suzanne Keen The Return of the Omniscient Narrator: Authorship and Authority in Twenty-First Century Fiction by Paul Dawson Feminist Narrative Ethics: Tacit Persuasion in Modernist Form by Katherine Saunders Nash Real Mysteries: Narrative and the Unknowable by H. Porter Abbott A Poetics of Unnatural Narrative edited by Jan Alber, Henrik Skov Nielsen, and Brian Richardson Narrative Discourse: Authors and Narrators in Literature, Film, and Art by Patrick Colm Hogan An Aesthetics of Narrative Performance: Transnational Theater, Literature, and Film in Contemporary Germany by Claudia Breger Literary Identification from Charlotte Brontë to Tsitsi Dangarembga by Laura Green Narrative Theory: Core Concepts and Critical Debates by David Herman, James Phelan and Peter J. Rabinowitz, Brian Richardson, and Robyn Warhol After Testimony: The Ethics and Aesthetics of Holocaust Narrative for the Future edited by Jakob Lothe, Susan Rubin Suleiman, and James Phelan The Vitality of Allegory: Figural Narrative in Modern and Contemporary Fiction by Gary Johnson Narrative Middles: Navigating the Nineteenth-Century British Novel edited by Caroline Levine and Mario Ortiz-Robles Fact, Fiction, and Form: Selected Essays by Ralph W. Rader edited by James Phelan and David H. Richter The Real, the True, and the Told: Postmodern Historical Narrative and the Ethics of Representation by Eric L. Berlatsky Franz Kafka: Narration, Rhetoric, and Reading edited by Jakob Lothe, Beatrice Sandberg, and Ronald Speirs Social Minds in the Novel by Alan Palmer Narrative Structures and the Language of the Self by Matthew Clark Imagining Minds: The Neuro-Aesthetics of Austen, Eliot, and Hardy by Kay Young Postclassical Narratology: Approaches and Analyses edited by Jan Alber and Monika Fludernik Techniques for Living: Fiction and Theory in the Work of Christine Brooke-Rose by Karen R. Lawrence Towards the Ethics of Form in Fiction: Narratives of Cultural Remission by Leona Toker Tabloid, Inc.: Crimes, Newspapers, Narratives by V. Penelope Pelizzon and Nancy M. West

Narrative Means, Lyric Ends: Temporality in the Nineteenth-Century British Long Poem by Monique R. Morgan Understanding Nationalism: On Narrative, Cognitive Science, and Identity by Patrick Colm Hogan Joseph Conrad: Voice, Sequence, History, Genre edited by Jakob Lothe, Jeremy Hawthorn, James Phelan The Rhetoric of Fictionality: Narrative Theory and the Idea of Fiction by Richard Walsh Experiencing Fiction: Judgments, Progressions, and the Rhetorical Theory of Narrative by James Phelan Unnatural Voices: Extreme Narration in Modern and Contemporary Fiction by Brian Richardson Narrative Causalities by Emma Kafalenos Why We Read Fiction: Theory of Mind and the Novel by Lisa Zunshine I Know That You Know That I Know: Narrating Subjects from Moll Flanders to Marnie by George Butte Bloodscripts: Writing the Violent Subject by Elana Gomel Surprised by Shame: Dostoevsky’s Liars and Narrative Exposure by Deborah A. Martinsen Having a Good Cry: Effeminate Feelings and Pop-Culture Forms by Robyn R. Warhol Politics, Persuasion, and Pragmatism: A Rhetoric of Feminist Utopian Fiction by Ellen Peel Telling Tales: Gender and Narrative Form in Victorian Literature and Culture by Elizabeth Langland Narrative Dynamics: Essays on Time, Plot, Closure, and Frames edited by Brian Richardson Breaking the Frame: Metalepsis and the Construction of the Subject by Debra Malina Invisible Author: Last Essays by Christine Brooke-Rose Ordinary Pleasures: Couples, Conversation, and Comedy by Kay Young Narratologies: New Perspectives on Narrative Analysis edited by David Herman Before Reading: Narrative Conventions and the Politics of Interpretation by Peter J. Rabinowitz Matters of Fact: Reading Nonfiction over the Edge by Daniel W. Lehman The Progress of Romance: Literary Historiography and the Gothic Novel by David H. Richter A Glance Beyond Doubt: Narration, Representation, Subjectivity by Shlomith Rimmon-Kenan Narrative as Rhetoric: Technique, Audiences, Ethics, Ideology by James Phelan Misreading Jane Eyre: A Postformalist Paradigm by Jerome Beaty Psychological Politics of the American Dream: The Commodification of Subjectivity in Twentieth-Century American Literature by Lois Tyson Understanding Narrative edited by James Phelan and Peter J. Rabinowitz Framing Anna Karenina: Tolstoy, the Woman Question, and the Victorian Novel by Amy Mandelker Gendered Interventions: Narrative Discourse in the Victorian Novel by Robyn R. Warhol Reading People, Reading Plots: Character, Progression, and the Interpretation of Narrative by James Phelan