beyond myth and metaphor- narrative in digital media

Beyond Myth and Metaphor: Narrative in Digital MediaRyan, Marie-Laure, 1946-

Poetics Today, Volume 23, Number 4, Winter 2002, pp. 581-609 (Article)

Published by Duke University Press

For additional information about this article

Access Provided by UNESP-Universidade Estabul Paulista Julio de Mesquita Filho at 04/08/11 11:26PM GMT

http://muse.jhu.edu/journals/poet/summary/v023/23.4ryan.html

Beyond Myth and Metaphor:Narrative in Digital Media

Marie-Laure Ryan

Abstract The concept of narrative has been widely invoked by theorists of digital tex-tuality, but the promotion of what is described as the storytelling power of the computerhas often relied on shallow metaphors, loose conceptions of narrative, and literarymodels that ignore the distinctive properties of the digital medium.Two myths havedominated this theorization. The myth of the Aleph (as I call it) presents the digitaltext as a nite text that contains an innite number of stories.The myth of the Holo-deck envisions digital narrative as a virtual environment in which the user becomesa character in a plot similar to those of Victorian novels or Shakespearean tragedies.Both of these myths rely on questionable assumptions: that any permutation of a col-lection of lexias results in a coherent story; that it is aesthetically desirable to be thehero of a story; and that digital narrativity should cover the same range of emotionalexperiences as literary narrative. Here I argue that digital narrative should emanci-pate itself from literary models. But I also view narrative as a universal structure thattranscends media.This article addresses the question of reconciling the inherent lin-earity of narrative structures with the multiple paths made possible by the interactivenature of the digital text by distinguishing four forms of interactivity, which resultfrom the cross-classication of two binaries: internal versus external interactivity;and exploratory versus ontological. Each of these categories is shown to favor dier-ent narrative themes and dierent variations of the universal narrative structure.

If we compare the eld of digital textuality to other areas of study in thehumanities, its most striking feature is the precedence of theory over theobject of study. Most of us read novels and see movies before we consult

Poetics Today : (Winter ). Copyright by the Porter Institute for Poetics andSemiotics.

tph

582 Poetics Today 23:4

literary criticism and cinema studies, but it seems safe to assume that avast majority of people read George Landows Hypertext . () beforethey read any work of hypertext ction. Or to take another example, weread full descriptions of what virtual reality technology would mean forour lives and for art long before VR became reality. (Even now its level ofsophistication is far from allowing the immersive experience promised byits early prophets.) In this article I would like to investigate one of the mostimportant forms that this advance theorizing of digital textuality has taken,namely, the use of narrative concepts to advertise present and future prod-uct. I will approach this topic in three ways: rst, through a critique of someof the (mis)uses of the concept of narrative in advertising and theoreticaldiscourse; second, through a taxomony of the various modes of user par-ticipations in digital narratives; and third, through a personal assessment ofthemost ecient way to exploit the resources of hypertext, themost literaryform of digital narrative.As a cognitive structure, narrative has such a grip on the mind that the

popular success of a genre ormedium involving language is crucially depen-dent on its ability to tell stories. It is because knowledge was encoded astales that it was eectively transmitted and remembered in oral societies;it is because of its narrative power that the novel emerged as the dominantliterary genre of the nineteenth and twentieth centuries; and it is because itgave new narrative dimensions to the novel and to the theater that cinemabecame the art of the twentieth century. In recent years, the concept ofnarrative has caught like re in cultural discourse, and the software industryhas duly followed suit by turning the metaphors of narrative interface andof the storytelling computer into advertising buzzwords. To take a coupleof recent examples of this free use of narrative terminology: Steven John-son concludes his popular book Interface Culture () with the pronounce-ment: Our interfaces are storieswe tell ourselves toward o senselessness;Steve Jobs, the founder and CEO of Apple, talks about the importanceof stories, of marrying technology and storytelling skills (Auletta : );and the package of a computer game named Starcraft advertises its capacityto let users tell their own stories. What this phrase really means is thatthe user can create new mazes and new levels of diculty, adding weaponsand characters to the game by selecting items from a xed repertory.To promote the narrative power of the computer, theorists of digital

media have either implicitly or explicitly relied on myths and metaphors.The rst part of this essay proposes a critique of three of these conceptual-izations: the metaphor of the narrative interface, the myth of the Aleph (alabel of my own making), and the myth of the Holodeck. The second part

Ryan Beyond Myth and Metaphor: Narrative in Digital Media 583

presents my own view of the role, and of the possible mode of realizationof narrative, in digital media.The terms myth and metaphor, in my view, are not synonymous, though

they present some overlap. Metaphor is generally dened as the transferof a concept from one domain to another. In this case it means the use ofnarrative conceptssuch as plot, storyteller, or characterto describe thedesign or mode of operation of a computer application whose purpose isnot in itself the telling of stories. By myth, on the other hand, I understanda theoretical model borrowed from ction that describes the artistic poten-tial of a digital form of narrative. This model is a myth not only becauseit is an imaginary construct but also because it oers an idealized repre-sentation of the genre it describes. My use of the term myth has both posi-tive and negative connotations: positive, because the pursuit of ideals is themost powerful of the forces that sustain art, and negative, because idealsare by denition not reachable.The myths I will evoke have served a usefulpurpose, since they energized the publics imagination, but they did so bysetting impossible or ill-conceived goals that raised false expectations.Thismay lead to a loss of interest in new media when these expectations turnout to be unfulllable.Before moving on withmy discussion, let me sketch some of my positions

on the nature of narrative:

Narrative is not coextensive with literature, ction, or the novel. Narrativity is independent of tellability. Narrative is not limited towritten or oral storytelling. It is amental rep-resentation that can be evoked bymanymedia andmany types of signs.

Narrativity is a matter of degree: postmodern novels are not nearly sonarrative as those of the nineteenth century.

As a mental representation, narrative consists of a world (setting),populated by individuals (characters), who participate in actions andhappenings (events, plot), through which they undergo change (tem-poral dimension).

1. The Metaphor of the Narrative Interface

The story of interface design, since the advent of the Macintosh, has beenshaped by a very simple and very powerful idea: computers are ugly, fear-some, inhuman, and they make people feel inadequate; it is thereforenecessary to hide them behind a metaphor that will make them pass forsomething else. The best-known of these disguises is the desktop conceptintroduced by Apple in the mid-eighties. This family of metaphors associ-


ates the modules of a software package with familiar objects, such as pens,les, folders, erasers, palettes, envelopes, and so on.These objects are repre-sented on the computer screen by visual icons that give the user a sense oftheir objectlike, physical presence. But the desktopmetaphor has the unfor-tunate side eect of limiting the computer to the role of a business machine.The next logical step in the concealment of the computers true nature istherefore the development of interface metaphors that suggest play andentertainment, even when the actual function of the software is the per-formance of professional tasks. All this explains the popularity of narrativemetaphors with software designers and Web page authors.The concept of narrative interface was introduced in a handful of

articles gathered in a collection edited by Brenda Laurel, The Art of Human-Computer Interface Design (). Some contributors understand narrative ina diegetic way, while others advocate a mimetic, or dramatic, applica-tion. The diegetic conception presupposes the existence of a narrator orstoryteller addressing an audience. If we apply this model to the case ofhuman/computer interaction, the computer will be the storyteller, and theuser will be the audience. In the mimetic or dramatic conception of narra-tive, by contrast, there is no need for a verbal act of narration, no need fora storyteller. Narrative comes into existence not by being told but by beingenacted. More precisely, it comes into existence in the mind of the spec-tator as an interpretation of what is seen or heard. All it takes in this caseto produce a narrative is agents who engage in an action that inspires theproper interpretation in the mind of the audience. In the mimetic model,the computer is not a storyteller but a character who interacts with the userin such a way that the user will regard their interaction as a story.The main proponent of the storyteller metaphor is Abbe Don ()

in her contribution to Laurels anthology, Narrative and the Interface.Though the article does not contain a systematic analysis of the features ofthe storyteller, it is easy to see what makes the metaphor so attractive tosoftware designers:

Oral storytelling is an inherently interactive situation.When a parenttells a story to a child, for instance, the child can ask questions or get theparent to expand some episodes (tell me again how Little Red RidingHood got out of the wolf s mouth). The story can be easily adaptedto the particular needs of the audience.

Storytelling, as a culturally universal phenomenon, provides modelsof understanding that people have already mastered. If software pro-grams exploit narrative structures, people will learn to use themeortlessly.


Though each storytelling performance is uniquely adapted to theneeds of the audience, the plot remains basically the same. Similarly,when users operate a program, they may have their own personalgoals, but a basic protocol and certain tasksmust be fullled every time.

Storytelling creates a sense of solidarity among the members of a com-munity. Operating a computer is a lonesome activity. The metaphorhelps overcome this loneliness by suggesting the supporting presenceof a user community.

The storyteller functions as the keeper and disseminator of culturalknowledge. Storytelling stands for the idea of teaching through enter-tainment. In oral cultures, information is traditionally embedded in aspellbinding narrative action.The Iliad, for instance, may have helpedmemorize the list of Greek tribes or the parts of a warriors equipment.The metaphor of the computer as storyteller means that the computeracts as a tutor and educator who knows how to turn the frustration oflearning the complicated operation of a program into a pleasurableexperience.

Storytelling provides an antidote to the cold indierence, rigid deter-minism and unbending logic of the computer. It gives a human faceto the machinethe face of compassionate computing.The metaphoralso gives a voice to a widespread nostalgia for an age when the tasksof everyday life could be performed through a set of tools whose func-tioning people could easily understand.

Yet there is a limit to the analogy. By casting the user as audience, thestoryteller metaphor ignores the dialogic nature of human/computer inter-action. Storytellingmay depend on the audience input, but it is essentially amonological form of discourse, with the storyteller monopolizing the oorfor a lengthy turn.The dramatic metaphors, with their implicit dialogism,seem therefore better suited to model the dynamics of human/computerinteraction. Their most inuential advocate is Brenda Laurel in her well-known book Computers as Theatre (). Her analogy can be decomposedinto a series of equivalencies:

The screen is a stage. The objects on the screen are the props. The user is a character who plays a role on the stage by manipulatingthe objects.

The interaction between the user and the objects produces a plot.

To be truly pleasurable to the user, this plot should be carefully scriptedby the system. Laurel recommends the adoption of an Aristotelian patternthat guides the user through the traditional dramatic curve of rise and fall in


tension. But she remains suspiciously vague on the question of the practicalimplementation of her concept of plot as well as on the kinds of applicationsthat lend themselves to a dramatic experience.How do these ideas of the storytelling computer and of the computer as

theater translate into concrete applications? At the present time the narra-tive metaphor has inspired two types of design. One involves the creationof a character and the other the creation of a setting.The best-knownmanifestation of the creation of a character is the Oce

Assistant of Microsoft Oce, a comic, friendly character selected by theuser from a menu of several choices that include a paper clip, a dog, a cat,and a cartoon version of Albert Einstein. What does this Einstein charac-ter do? He rolls his eyes, aps his arms, shakes his head, gets mad, quietsdown, nods approvingly. He remains active as long as a time-consumingtask is going onsuch as saving, opening, copying, or downloading a le.Through his cute antics, the Oce Assistant fullls several functions, somedramatic, some communicative, and some more narrowly narrative. Themost obviously dramatic one is the Shakespearean function of comic relief.Even when Einstein has nothing substantial to say, he alleviates the serious-ness of the task performed by the user.The danger of comic relief, of course,is that the routine is funny only the rst time around; most of the people Ihave asked about the Oce Assistant nd him a distracting nuisance. Thesecond function of Einsteina communicative rather than strictly narra-tive oneis what Roman Jakobson () has called the phatic function.This function consists in maintaining contact with an interlocutor to makesure that the channel of communication is still open, for instance, by pipingmuzak over the telephone line while putting a customer on hold.Throughhis little dance, Einstein entertains users and tells them that the system is notdead when control has been taken over by the machine for a lengthy periodof time. The third, and most specically narrative, function of Einstein isthe helper function. If the user tries to close a le without saving it rst, Ein-stein gets upset and asks: Do you really want to lose your changes? If theuser types Dear Susie, Einstein pops up in the corner and says, It lookslike you want to write a letter. Do you want me to help? Though a singlecharacter does not make a story, the gure of the Oce Assistant suggestsan implicit scenario that puts the user in the role of a hero of a Proppianfairy tale: as the user-hero I have a task to accomplish; in order to do so, Imust tame a mean machine; but along the way, I meet a friendly characterwho helps me conquer the villainous system and fulll the mission.The other concretization of the concept of narrative interface is the

design of a setting that encourages make-believe. This type of metaphorplaces users in a role that conceals what they are actually doing and dictates


a scenario that gives them a sense of purposeful action. My example of nar-rative setting is the movie-making metaphor of the program Director. Asthe name of the program suggests, the user is cast as a lmmaker, the screenis the stage (the metaphor is somewhat mixed between lm and theater),the objects that the user puts on the stage are the members of the cast, theybehave according to scripts written by the director, the pairing of scriptsand cast members is called the score, and the end product is called a movie.As a Director user, I can personally attest that I experienced an exhilarat-ing feeling of power at the thought that I had become a lmmaker, able tosummon actors on the stage and to dictate their behavior. The metaphoralso greatly facilitated my understanding of the various types of objects Iwas dealing with in the program.Those who expect from the term narrative interface a spellbinding plot with

lively characters and surprising twists will be deeply disappointed by theserather trivial scripts and supercial analogies. But it is precisely the banalityof the narrative scenario that makes it ecient. In the design of software,narrative is not an end in itself but a means toward a goal, and this goal isto facilitate the operation of the program. Interface metaphors, not unlikepoetic ones, fulll their rhetorical and pedagogical functions by relating astrange new world to a familiar one.

2. Hypertext, and the Myth of the Aleph

Myth of the Aleph is my way of describing how the early theorists ofhypertext conceived the narrative power of the new type of text. The termcomes from a short story by Jorge Luis Borges (), in which the scrutinyof a cabbalistic symbol enables the experiencer to contemplate the wholeof history and of reality, down to its most minute details. The Aleph is asmall, bound object that expands into an innity of spectacles.The experi-encer could therefore devote a lifetime to its contemplation. Though theydid not explicitly invoke the model of the Aleph, the pioneers of hyper-text theory conceived the new literary genre in strikingly similar terms.For theorists such as George P. Landow, Jay David Bolter, and MichaelJoyce, hypertext is a textual object that appears bigger than it is becausereaders could spend hoursideally, their entire lifetimesunraveling newstories from it. As Michael Joyce (: ) puts it: Every reading . . .becomes a new text. . . . Hypertext narratives become virtual storytellers.Likemany authors before themProust,Mallarm, James Joycethe pio-neers of hypertext dreamed their brainchild as the ultimate literary work,the sum of all possible narratives, the only text the reader will ever needbecause its meaning cannot be exhausted.


This conception of hypertext as amatrix that contains an innite numberof narratives is particularly prominent in the work of George Landow. Oneof the chapters of his seminal book Hypertext . () is titled Recong-uring Narrative. Since the word narrative is ambiguous between narrativediscourse and mental representation (the technical story concept of nar-ratologists, in its opposition to discourse), Landows claim can be under-stood in two ways. The rst is the discourse sense: hypertext changes theway narrative structures are encoded, how they come to the reader, howthey are experienced in their dynamic unfolding. The feature that enableshypertext to recongure narrative on the discourse level is, evidently, itsinteractive design, a design that enables the reader to select many dierentpaths through the narrative discourse and to view its units inmany dierentorders. But this newway of presenting stories does not mean that the storiesthemselves are radically dierent from traditional narrative patterns.Therecould be one xed story that comes to the reader in many dierent ways,depending on which path is chosen through the network.But this rather tame interpretation of reconguring narrative is not

what Landow (: ) has in mind: Hypertext, which challenges nar-rative and all literary form based on linearity, calls into question ideas ofplot and story current since Aristotle. TheAristotelian ideas that hypertextchallenges are:

() xed sequence, () denite beginning and ending, () a storys certain de-nite magnitude, and () the conception of unity and wholeness associated withall these other concepts. In hyperction, therefore, one can expect individual forms,such as plot, characterization, and setting, to change, as will genres or literary kindsproduced by congeries of these techniques. (Ibid.: ; italics added)

When he writes that plot changes, does Landow mean that in hypertextplot becomes something entirely dierent: a representation in which eventsfreely oat in time rather than forming a sequence at all? Or does he meanthat every reading resequences events, and that plot consequently changesall the time, but within reasonably constant parameters? The rst interpre-tation alters the concept of plot beyond recognition: if events oat freelyin time, without forming a xed sequence, time itself disappears, since oursense of time is tied to a sense that moments and events succeed each otherin an inexorably linear fashion. Without sequence, moreover, there is nocausality and no logical coherence. How can one still speak of plot and ofstory under these conditions?Let me therefore consider the other option: hypertext challenges the

notion that there is only one sequence and one plot in the text and thatreaders are done when they have reconstructed an event trajectory that


leads from a beginning to an end. It is up to the reader to rationalize everyreading, and consequently every arrangement of lexias, into a coherentstory. In a hypertext environment a lack of linearity does not destroy nar-rative. In fact, since readers always, but particularly in this environment,fabricate their own structures, sequences, or meanings, they have surpris-ingly little trouble reading a story or reading for a story (ibid.: ). In thisinterpretation, every traversal yields a possible story, in the semantic sense,because it is the reader who constructs the story out of the textual segments.Hypertext is like a construction kit: it throws lexias at its readers, one at atime, and tells them: make a story with this.Landow (ibid.: ) compares this situation to the mental activity of the

speaker of a language who forms an innite number of sentences out ofnite grammar: As readers we nd ourselves forced to fabricate a wholestory out of separate parts. . . . It forces us to recognize that the activeauthor-reader fabricates text andmeaning from anothers text in the sameway that each speaker constructs individual sentences and entire discoursesfrom anothers grammar, vocabulary, and syntax. I nd this analogy fal-lacious, because it hides an important dierence: the linguistic competenceof the speaker is an internalized knowledge of the syntactic rules and lexi-con of a language. To make a sentence, the speaker selects patterns andwords from a knowledge base more or less completely available to the mindbefore the speaker begins the sentence. But in hypertext, lexias come one ata time; and the reader must create a story on the move, without knowingwhat lexia will come next.Another problem with Landows suggestion is that if we take literally

the claim that every traversal of the database produces a dierent story, areader who encounters three segments in the order A then B then C willconstruct a dierent story than a reader who encounters the same segmentsin the order B then A then C. It is only if sequence plays a crucial role indetermining meaning that hypertext can be viewed as an Aleph that con-tains potentially a large number of dierent stories. If the reader could placeA, B, and C wherever narrative logic asks for in a developing narrative pat-tern, it would not matter in which order A, B, and C are encountered.Thisemphasis on themeaningfulness of sequence hits, however, a serious logicalobstacle. Textual fragments are like the pieces of a jigsaw puzzle; some teasily together, and some others do not because of their intrinsic content,the narrative equivalent of shape.The reader will admittedly do whatWolf-gang Iser () calls ll in the blanks to construct a plot, that is, imagineuntold episodes that glue the lexias together; but when we deal with a typeof meaning as narrowly constrained as narrative, lling in the blanks hasits limits. It is simply not possible to construct a coherent story out of every


permutation of a set of textual fragments, because fragments are implicitlyordered by relations, such as logical presupposition, material causality, andtemporal sequence. What, for instance, will I do if, in the course of myreading, I encounter a segment that describes the death of a character and,later on, a segment that describes the actions of this character when alive?Should I opt for a supernatural interpretation, according to which the char-acter was resurrected? This may be appropriate in certain contexts, but ifthe text has so far created a realistic, everyday type of world, the theme ofresurrection would threaten its thematic coherence. Most readers, accus-tomed to techniques of ashback and ash-forward, will construct a mentalimage in which being alive precedes being dead; but if such rearrangementoccurs, readers will reconstruct the same story regardless of the order inwhich they read lexias.If it is unrealistic to expect that readers will be both willing and able to

provide missing links to connect segments in a narratively meaningful wayfor every dierent order of appearance, we must abandon the Alephic con-ception of a new storywith every reading session and replace it with amodelthat describes the readers activity as the arrangement of textual segmentsinto a global pattern that slowly takes shape in the mind.This model is thejigsaw puzzle. Just as we can work for a time on a puzzle, leave it, and comeback to it later, readers of hypertext do not start a new story from scratchevery time they open the program but, rather, construe a global representa-tion over many sessions, completing or amending the picture put togetherso far.

3. Virtual Reality Narrative, and the Myth of the Holodeck

My second myth, the Holodeck, has been proposed by theorists as a modelof what narrative could become in amultisensory, three-dimensional, inter-active virtual environment. Its main proponent is JanetMurray in her well-known book Hamlet on the Holodeck (). But the concept of the Holodeckhas also been invoked by Jaron Lanier, the visionary developer of VR tech-nology (mentioned in Ditlea ) and by Michael Heim (), its no lessvisionary theorist.All of these writers borrow theHolodeck idea from the popularTV series

StarTrek.TheHolodeck is a kind of VR cave, to which the crewmembers ofthe starship Voyager retreat for relaxation and entertainment. In this cave, acomputer runs a three-dimensional simulation of a ctional world, and thevisitorthe interactorbecomes a character in a digital novel.The plotof this novel is generated live, through the interaction between the humanparticipant and the computer-created virtual characters. As Murray (:


) writes: The result is an illusory world that can be stopped, started, orturned o at will but that looks and behaves like the actual world. . . . TheStar Trek Holodeck is a universal fantasy machine . . . a vision of the com-puter as a kind of storytelling genie in the lamp. It enables crew membersto enter richly detailed worlds . . . in order to participate in stories thatchange around them in response to their actions.The rst chapter of Murrays book describes a Star Trek episode in which

Kathryn Janeway, the female commander of the starship Voyager, sneaksinto the Holodeck and becomes Lucy, the governess of the children in anaristocratic Victorian household. Lucy falls in love with the father of thechildren, Lord Burley, and they exchange passionate kisses, but the veryresponsible Kathryn realizes that this love for a virtual human is detrimen-tal to the fulllment of her duties in the real world, and she eventuallyorders the computer to delete the character. Murray (ibid.: ) interpretsthis action as evidence that VRbased interactive drama can match boththe entertainment and the educational value of literary narrative: TheHolodeck, like any literary experience, is potentially valuable in exactlythis way. It provides a safe place in which to confront disturbing feelingswe would otherwise suppress; it allows us to recognize our most threaten-ing fantasies without becoming paralyzed by them. Some readers may bepuzzled by this use of a science-ctional scenario tomake predictions aboutthe artistic potential of what is supposed to become someday a real tech-nology. Even more strangely, however, Murray bases her assessment of thewholesome eect of Holodeck narrative on an episode in which, precisely,the heroine is forced to shut down the system in order to be able to func-tion in the real world. It is as if literary narrative were only good for thosereaders who throw the book away midway though their reading.The viability of the concept of the Holodeck as model of digital narra-

tive is questionable for a number of reasons: technological, algorithmic, butabove all psychological. These problems are evident in Murrays descrip-tion of how she envisions aVRbased interactive version of themovieCasa-blanca. The object of this version would be to oer the interactor to havedierent adventures by assuming the roles of several distinct characters, allof whom are pursuing their own destinies in the French-controlled colo-nial city during World War II (ibid.: ). We can ignore the question ofthe creation of the setting, since it is primarily a problem of hardware andgraphic design.The truly problematic issues are those of plot and dialogue.Unlike most hypertext theorists, Murray has a fairly strict idea of plot, andshe is not willing to leave responsibility for its creation to the interactor orto chance. In her vision of the Holodeck, plot and dialogue are controlledby the author and users step into the roles of predened characters rather


than creating their own.Murray envisions the plot as a branched or multi-form designmade ofmany prerecorded unitsepisodes, themes, or mor-phemes, asMurray (ibid.: ) calls them in reference to ProppsMorpholoof the Folktale ( [])that are activated by the system in response tothe interactors actions. The many branches in the plot correspond to thevarious decisions the interactor could make in the situations presented bythe system. But how could the interactor retain a reasonable freedom ofaction throughout the performance without taking the plot in a directionfor which there is no ready-made, logically coherent response stored in thesystem? And howwould the system handle dialogue between the interactorand the characters? Only an impractically large amount of author-created,prerecorded dialogue could allow the system to produce meaningful con-versation between the interactor and the virtual characters.Thismeans thatthere will be no dialogue or that the users freedom of speech will be lim-ited to something of the scope of yes/no responses. Or perhaps even thatthe system will take full control over dialogues and generate the words ofthe interactor.But even if all these problems could be resolved, even if the right bal-

ance could be struck between user freedom and system control, even if thesystemmanaged to coax, rather than coerce, the interactor to take dramati-cally optimal paths, an important question remains. What kind of grati-cation will the experiencer receive from becoming a character in a story?It is important to remember at this point that, even though interactors areagents and in this sense coproducers of the plot, they are above all the bene-ciaries of the performance. The entertainment value of the experiencedepends on how the interactors relate to their avatars. Will interactors belike actors playing a role, innerly distanced from their characters and simu-lating emotions they do not really have, or will interactors experience theircharacter as their own self, actually feeling the love, hate, fears, and hopesthat motivate the characters behavior or the exhilaration, triumph, pride,melancholy, guilt, or despair that may result from the characters actions?If we derive aesthetic pleasure from the tragic fates of literary charac-

ters such as Anna Karenina, Hamlet, or Madame Bovary, if we cry forthem and fully enjoy our tears, it is because our participation in the plotis a compromise between identication with the character and distancedobservation. We simulate mentally the inner life of each character, wetransport ourselves in imagination into the mind of each, but we remain,at the same time, conscious of being external witnesses. But in the Star TrekHolodeck, which is of course an imaginary construct, the interactor experi-ences emotions from the inside, to use a concept proposed by KendallWalton (: ). This is why Kathryn Janeway must delete the char-


acter of Lord Burley: even though she experienced the fairy-tale romanceof a poor governess conquering the love of a handsome aristocrat, her lovefor him threatened her ability to fulll her real-world mission as well as herrelationship to her real-world boyfriend. Had she taken another course ofaction, she might have run into a much less pleasant scenario: loving LordBurley but being rejected, remaining cold to his passion and regarding hisadvances as a nuisance, or experiencing nothing more than a strictly busi-nesslike servant-master relationship. Interactors would have to be out oftheir mindsliterally and gurativelyto want to live these plots in therst-personmode. (By rst- and third-personmode I do notmean the tradi-tional narrative voices but whatWalton calls experiencing from the insideand from the outside.)Yet if becoming a ctional character will often result in a rather

unpleasant experience, Murray does not place restrictions on the kind ofstories that will be suitable for Holodeck-style enactment. She takes greatpains to demonstrate that new media can express the entire spectrum ofhuman emotions. Digital narratives will be as good for gut-wrenchingdrama as for action plots. In her imaginary Casablanca simulation, forinstance, she would like to oer the user the choice of causing the deathof another character by denouncing him to the Nazis. To rival the ethicaldimension of literary narrative, the simulation should make the user con-scious of the moral consequences of this action.This means inicting guilton the despicable interactor.Murray (: ) imagines the following end-ing for the traitor scenario: You could nd yourself sitting at a table with anewspaper report of the death lying in front of you next to a bottle and glass.You would be able to pour the liquor and raise the glass but not get up fromthe table. This enforced immobility would suggest the despair of a personabout to drink himself to death. This is a very moral ending indeedbutwill interactors actually feel guilt, or will they respond by thinking: Thepoor devil feels guilty, and he sure deserves it. This reaction would meanthat users relate to their avatars in the third-person mode.The user will bemore a puppet master who makes decisions for a certain character, moti-vated by the curiosity of nding out what will result from these decisions,than a human being existentially, emotionally, andmorally caught in a cer-tain situation. By maintaining a safe distance between reader and charac-ters, literature has been able to explore the whole spectrum of human emo-tions without inicting intolerable suering on the reader. Any attempt toturn empathy, which relies on mental simulation, into emotions felt fromthe inside would in the vast majority of cases cross the fragile boundarythat separates pleasure from pain.Only selected types of emotional experiences, and consequently selected


types of plots, lend themselves to a rst-person perspective. If we con-sider the whole gamut of ctional characters, which ones would we reallylike to impersonate? Given the choice, would we identify with somebodylike Hamlet, Emma Bovary, Gregor Samsa in The Metamorphosis, Oedi-pus, Anna Karenina, and Brutus in Julius Caesar, or would we rather enterthe skin of the dragon-slaying hero of Russian fairy tales, Alice in Won-derland, Harry Potter, and Sherlock Holmes? If we pick a character fromthe second list, this means that we prefer becoming a rather at characterwhose involvement in the plot is not aective but a matter of exploring aworld, solving problems, performing actions, competing against enemies,and above all dealing with objects in a concrete environment.This kind ofinvolvement is much closer to playing a computer game than to living aVictorian novel or a Shakespearean drama.Toward the end of her book Hamlet on the Holodeck, Murray (ibid.: )

writes rather cryptically: Narrative beauty is independent of medium.This statement can be interpreted in two ways, one that I nd profoundlytrue, and the other profoundly false. The false interpretation claims thatsince narrativity is a cognitive pattern or mental representation indepen-dent of medium, all media are equally equipped to represent a given plot.This means that in some distant and very questionable future, when AI issuciently advanced to generate coherent plots in response to the usersaction, and to do this in real time, we will have an interactive version ofHamlet, or one of any other imaginable plot. Digital media will oer anenhanced version of literary classics, and theywill truly become the art formof the twenty-rst century.This interpretation not only ignores the idiosyn-cratic features of each medium, it also assumes rather presumptuously thatwhat digital technology adds to existing media is necessarily a dimensionthat enhances narrativity.The other interpretation, the one that I endorse,says that the abstract cognitive structure we call narrative is such that it canbe called to mind by many dierent media, but each medium has dierentexpressive resources and will therefore produce a dierent concrete mani-festation of this general structure. Put in simpler words: there are plot typesand character types that are best for the novel, others are best for oral story-telling, and yet others are best for the stage or the cinema. The question,then, is to decide which types of stories are suitable for digital media.

4. Narrativity and Interactivity

The answer to this question is crucially dependent on what constitutes thetruly distinctive resource of digital media, namely, the ability to respond tochanging conditions in the global state of the computer.When the change


in conditions is determined by the users input, we call this resource inter-activity. By singling out this one feature, it may seem that I am neglectingothers, such as the realism and uidity of digital images, the sense of spacethat comes from navigating a virtual world, the dramatization of time thatoccurs when players compete against moving objects in computer games;or the immersive nature of virtual worlds. But all these features can betraced back to the responsiveness of the system to the actions of the user.It is because the display adapts itself to the position of the cursor, whichstands for the body of the user, that digital environments convey the experi-ence of movement; it is in turn the experience of movement that leads toa heightened sense of time, of space, and of the presence of the environ-ment.The interactive nature of digital worlds is the true foundation of theirimmersivity (Ryan ).For the purpose of my argument I would like to distinguish four strategic

forms of interactivity on the basis of two binary pairs: internal/external andexploratory/ontological.These two pairs are adapted fromEspen Aarseths(: ) typology of user functions and perspectives in cybertexts,which is itself part of a broader cybertext typology. But I use dierent labelsthat shift the emphasis toward the users relation to the virtual world. Thepoint ofmy discussion of these categories, however, is not to revise Aarsethstypology, but to show howdierent types of interactivity open dierent pos-sibilities on the level of narrative themes and plot conguration.

4.1. Internal/External InteractivityIn the internal mode, users project themselves as members of the ctionalworld, either by identifying with an avatar or by apprehending the virtualworld froma rst-person perspective. In the externalmode, users are situatedoutside the virtual world.They either play the role of a godwho controls thectional world from above or they conceptualize their own activity as navi-gating a database. This distinction is a matter of degree: there are digitaltexts that situate the user at a variable distance with respect to the ctionalworld or that locate the user at the periphery, not quite in, not quite out.The dichotomy internal/external corresponds roughly to Aarseths

(: ) distinction between personal and impersonal perspective: aworld-internal participation will logically result in the users personica-tion, since worlds are spaces populated by individuated existents, whileworld-external involvement does not require a concrete persona.The onlypotential dierence between Aarseths labels and mine is the case of a userwho is projected as a powerful gure external to the playing eld and whomakes strategic decisions for the participants, such as the commander inchief of an army, a sports coach, an author writing a novel, or a specic god.


4.2. Exploratory/Ontological InteractivityIn the exploratory mode, users navigate the display, move to new observa-tion points, alter their perspective, or examine new objects in order to learnmore about the virtual world. But this activity does not make ctional his-tory, nor does it alter the plot; users have no impact on the destiny ofthe virtual world. In the ontological mode, by contrast, the decisions of theusers send the history of the virtual world on dierent forking paths.Thesedecisions are ontological in the sense that they determine which pos-sible world, and consequently which story, will develop from the situationin which the choice presents itself. This distinction is much more strictlybinary than the preceding one: the user either does, or does not, have thepower to intervene in the aairs of the ctional world.In his own taxonomy, Aarseth (: ) comes up with two roughly simi-

lar categories, but his exploratory and congurative are part of a longerlist of user functions that also comprises the opposition textonic andinterpretive: textonic means the ability to add permanent elements to thetext; interpretive, the lack of this feature. I view these last two categories asdierent ways to fulll the exploratory and ontological functions.Whereasthe merely interpretive stance is compatible with both an exploratory andan ontological involvement, textonic participation presupposes ontologicalinvolvement, since the text added by the user contributes to the shaping ofthe ctional world. My dichotomy also bears some resemblance to BrianMcHales (: ) distinction between an epistemological dimension,dominant in modernist literature, and an ontological one, dominant inthe postmodernist era. Exploratory interactivity is clearly dictated by epis-temological concerns, since its purpose is to learn more about the ctionalworld. But my ontological category is far less metaphysical than the lit-erary dominant described by McHale. It is a largely nonreexive way ofperforming world-creating actions rather than a questioning of the natureof being.The cross-classication of the two binaries leads to four combinations.

Each of them is characteristic of dierent genres and aords dierent nar-rative possibilities.

Group 1: External-exploratory interactivity. In the texts of this groupmostly classical hypertexts, such as the novels of Michael Joyce, StuartMoulthrop, orMark Amerikathe user is external to both the time and thespace of the ctional world. Interactivity resides in the freedom to chooseroutes across a textual space, but this space has nothing to dowith the physi-cal space of a narrative setting. The implicit map of the text represents a


network of lexias, not the geography of a ctional world. In classical hyper-text, the network is usually too densely connected for the author to con-trol the readers progression over signicant stretches. Randomness sets inafter one or two transitions. But randomness is incompatible with the logi-cal structure of narrative. Since it would be impossible for the author toforesee a coherent narrative development for each path of navigation, theorder of discovery of the lexia cannot be regarded as constitutive of nar-rative sequence. The only way to preserve narrative coherence under suchconditions is to regard the text as a scrambled story that the reader putsback together, one lexia at a time.This type of interactivity is external, because the text does not cast the

reader as a member of the ctional world. Readers regard the text less as aworld in which to immerse themselves than as a database to be searched oras a construction kit for assembling aworld. If we conceptualize the text as apuzzle, interactivity is exploratory, because the readers path of navigationaects not the narrative events themselves but only the way in which theglobal narrative pattern (if there is one at all) emerges in the mind. Simi-larly, with a jigsaw puzzle, the dynamics of the discovery dier for everyplayer, but they do not aect the structure that is put together. Moreover,just as the jigsaw puzzle subordinates the image to the construction pro-cess, external-exploratory interactivity de-emphasizes the narrative itselfin favor of the game of its discovery. This mode is therefore better suitedfor self-referential ction than for narrative worlds that hold us under theirspells for the sake of what happens in them. It promotes a metactionalstance, at the expense of immersion in the ctional world.This explains whyso many literary hypertexts oer a collage of literary theory and narrativefragments.

Group 2: Internal-exploratory interactivity. In the texts of this category, theuser takes a virtual body with her into the ctional worldto paraphraseBrenda Laurel (: )but her role in this world is limited to actionsthat have no bearing on the narrative events. She inhabits the space of thectional world but not the time of the narrative events. (I am using the femi-nine form because it is through texts of this type that the game industry istrying to reach a female audience.)1 The user has a seat on the stage; shemay even play an active role, such as that of a traveler, an explorer, a his-torian, or a detective who tries to solve a mystery, but she is not the hero ofthe action. In the words of Thomas Pavel (: ), she is a non-votingmember of the ctional world. The user exercises her agency by moving

. On this topic, see Cassell and Jenkins .


around the ctional world, picking up objects and looking at them, viewingthe action from dierent points of view, investigating a case, and trying toreconstitute events that have taken place a long time ago.This type of interactivity lends itself to several types of plot:

The mystery plot, in which two narrative levels are connected: oneconstituted by the actions of the detective, the other by the story tobe reconstructed. In this conguration, the second level is predeter-mined, while the rst is created in real time by the actions of the user.Example: the computer game Myst, in which the user explores anisland and solves certain puzzles to crack themystery ofwhat happenedin the past.

The parallel plot, or soap opera type, in which a large cast of char-acters acts simultaneously in dierent locations, so that it is necessaryfor the user to move from one location to another to another to followevery thread in the plot. Example: the nowdefunct Internet soap operaThe Spot, which followed the intersecting destinies of several charac-ters. New episodes were posted everyday, each written from the pointof view of one of the characters. The user could follow one characterfor a while, then switch to another. She could look at their letters anddiaries, and in a possible variation, she could access a version of thestory told in the third person.

The spatial narrative,whosemain theme is travel and exploration.Thiscould be an electronic version of Alice inWonderland, where Alice wouldnot really do anything but rather stumble into the lives of the othercharacters and observe them for a while.

The narrative of place, which is a combination of parallel plot andspatial narrative.The purpose of the narrative of place is not to travelacross vast expanses, as does the narrative of space, but rather, toexplore in depth a specic location, to look at all the objects containedin it, and to meet all of its inhabitants. An example of this type isthe hypertext ction Marble Springs () by Deena Larsen, a text thatinvites the reader to explore the map of a Colorado ghost town andtells, in short poems, about the lives of its female inhabitants. (The livesof the men are left to the reader to write.) In the narrative of place,interest resides not in an overarching plot, that is to say, not in a grandnarrative of the macro level, but in the little stories that the userdiscovers in all the nooks and crannies of the ctional world.

Group 3: External-ontological interactivity. Here the user is like the omnip-otent god of the system. Holding the strings of the characters, from a posi-


tion external to both the time and the space of the ctional world, theuser species their properties, makes decisions for them, throws obstaclesin their way, and creates dierent destinies for them by altering their envi-ronment. A classical example of this type of interactivity is the interactiveDVD movie Im Your Man (, directed by Bob Bejean. A Choice PointFilm).The movie involves three characters: a villain, Richard; a fool, Jack;and a good girl, Leslie. At one of the branching points, the movie asks thespectator if Richard should kill Leslie or seduce her. At another point, thespectator faces the choice of making Jack act like a hero or a coward. Bymaking a decision, the spectator assumes an authorial stance toward theprotagonists, since the choices aect their moral characters, which in turndetermines their fates. This activity of playing with parameters to see howthe system will evolve is equivalent to the operation of a simulation system.Since the operator of the narrative system is external to the ctional world,he or she has no strong interest at stake in any particular branch of its vir-tual history; gratication resides instead in the contemplation of the wholeeld of possibilities. The individual forking paths in the plot are thereforeless interesting than the global system of their interconnections.From a thematic point of view, this mode of interactivity lends itself to

what I would call, following Niall Ferguson (), virtual history nar-ratives. In the newly fashionable eld of virtual history, serious scholarsdebate such questions asto plagiarize Pascalwhat would have beenthe fate of the world if Cleopatras nose had been shorter. Themeaningful-ness of such exercises is rooted in the belief that destiny is governed by smallrandom events that lead to large-scale dierences, if the system is allowedto run its course without further intervention for a long period of time.The same idea underlies the so-called buttery principle of chaos theory:a buttery apping its wings in Beijing aects the weather in Corsica.The combination of ontological and external interactivity would be illus-

trated by the conception of hypertext as an Aleph and of the reader ascoauthor of the plot, if indeed it were possible to nd narrative coherencein each particular traversal of a hypertextual network. But as I have alreadysuggested, narrative coherence is impossible to maintain in a truly complexsystem of links.We need therefore simpler structures, structures with fewerbranches and fewer decision points, so that every path can be individuallydesigned by the author. Once the user has made a choice, the narrativeshould be able to roll by itself for an extended period of time; otherwise,the system would lead to a combinatory explosionor fall back into ran-domness, the deathbed of narrative coherence.The best-known example of a narrative system with an ontological-

external type of interactivity is the series of childrens books Choose Your


Own Adventure.The underlying structure of these stories is a fairly simpletree-shaped diagram, on which each branch is kept separate from theothers.This enables the designer to maintain a strict control over the linearsequence of events.2

Another example of external ontological interactivity is the simulationgame, such as Simcity, Simlife, Caesar, orThe Sims. In these games, playersrule over a complex system, such as a city, an ant colony, an empire, or afamily, and their decisions aect the evolution of the system. Even in a gamelike The Sims, where the player creates a number of individuals, the maincharacter in the developing narrative is a collective entity, and this char-acter has no consciousness of its own: it is just the sum of multiple micro-processes. The range of possible developments at any given point dependson the possibilities of action oered by the various objects and individualswithin the ctional world. For instance, a computer in The Sims aordstwo types of action: play games or look for a job.The choice of one of theseaordances aects the life and the options of several members of the c-tional world; for instance, if the user decides that Betty inThe Sims will usethe computer to get a job, Betty will earn money, and she will be able tobuy a wider variety of commodities. This in turn may aect Bobs feelingsfor Betty. The possibilities of action evolve during the run of the program,and since aordances are determined by the global state of the system, aswell as by the nature of the objects, the users choices will always producea coherent narrative development.While the operation of a simulation system requires a godlike position of

power, many of the games mentioned above try to increase dramatic inter-est by casting the user as a member of the ctional world. In Caesar, forinstance, the user is the ruler of theRomanEmpire; in Simcity, themayor ofthe city.Themayor or the emperor are external interactors, because they donot exist in the same space and time as their subjects.They rule the systemfrom above, as the gods eye perspective of the graphic display indicates,and they do not operate in a simulacrum of real time, since they have all thetime in the world to make their decisions. But they are also internal partici-pants, because their personal fates are at stake in the way they govern.Themayor will be voted out of oce if his or her administration of the city does

. The second-person form should not be taken to mean that the reader is internalized ascharacter; the texts of the series are usually told in the third person. Even when the text usesthe second person, the reader relates to this you as if the reader were a he or a she.In a branching story about Pinocchio, for instance, the reader holds the strings of a puppetnamed Pinocchio and maintains an authorial perspective over the plot that diminishes thereaders emotional involvement in the current destiny of Pinocchio. A sane reader will notfeel crushed if his or her decisions lead Pinocchio to be turned into a donkey or swallowedby a whale: there will always be another run of the system, another destiny to be explored.


not please his or her constituents, and Caesar will be dethroned if the Bar-barians invade his empire. This combination of features places the gamesin question halfway between categories and .

Group 4: Internal-ontological interactivity. If the Holodeck could be fullyimplemented, this is where it would belong. In the meantime, the categorywill have to be represented by computer games of the action and adven-ture type, such as Doom, Quake, or Half-Life. Here the players are cast ascharacters situated in both the time and the space of the ctional world.The actions of the players determine the fate of their character (avatar in thetechnical jargon) and, by extension, the fate of the ctional world. Everyrun of the system produces a new life and consequently a new life story forthe avatar. This narrative is created dramatically, by being enacted, ratherthan diegetically, by being narrated.In this type of system, interactivitymust be intense, since we live our lives

by constantly engaging with the surrounding world. Most players are toodeeply absorbed in the pursuit of a goal to reect on the plot that they writethrough their actions, but when people describe their sessions with com-puter games, their reports typically take the form of a story. Consider, forinstance, this review by Peter Olafson () of the game Combat Mission,which simulates the German campaign into Russia during World War II:

My two panzer IVG tanks got lucky. Approaching the crossroads, they cleareda rise and caught two Sherman tanks out of position, one obstructing the aimof the other. Concentrating their re, they quickly took out the Allied units andthe surviving crews abandoned the aming hulks and retreated into the woodsnearby.

After three paragraphs of such prose, the account of the session concludeswith:

The computer commander knew it was licked. It began to pull back, and I nallyallowed myself to breathe again. Its the rst war game I can recall in which Iveresponded emotionally to a victory, and I knowwhy. It felt as though I was there.(Ibid.)

As this retelling demonstrates, the narrativity of the action game lies in thetrace of the actions performed by the player.Many people will rightly argue that creating a narrative is not the point

of adventure/action games.Computer games are played for the sake of solv-ing problems and defeating opponents, of rening strategic skills, and ofparticipating in on-line communities, not for the purpose of creating a tracethat reads as a story. Few people in their right minds will bother to recorda game session and watch the replay as a movie. (Developers may do that,


but for a dierent purpose.) The drama of the game is only worth experi-encing as an active participant; it is meant to be lived and not spectated.Yetif narrativity were totally irrelevant to the enjoyment of games, why woulddesigners put so much eort into the creation of a narrative interface?Whywould the graphics be so sophisticated? Why would the task of the playerbe presented as ghting terrorists or saving the earth from invasion by evilcreatures from outer space, rather than as gathering points by hittingmov-ing targets with a cursor controlled by a joystick? The evolution of com-puter games, since the early days of PacMan, Tetris, or Paddle Ball, hasbeen toward greater visual realism, which alsomeans toward greater narra-tivity, since in at least one of its denitions, realism is the power to constructa coherent, believable world that functions as setting for a dramatic action.The narrativity of action games functions as what Kendall Walton (:) would call a prop in a game of make-believe. It may not be the raisondtre of games, but it plays such an important role as a stimulant for theimagination that many recent games use lengthy lm clips, during whichthe player can only watch, to enrich the plot. (The fact that it is necessary totemporarily remove control from the user to establish the narrative frameis a further indication that interactivity is not a feature that facilitates theconstruction of narrative meaning.)At present, the thematic and structural repertory of ontological/internal

interactivity is quite limited. Adventure and role-playing games3 imple-ment the archetypal plot that has been described by JosephCampbell ([]) andVladimir Propp ( []): the quest of the hero across a landlled with many dangers to defeat evil forces and gain a desirable object.Some plots, however, deviate from the archetype in two ways: the hero canlose, and the adventure never ends. In most action games, the archetype isfurther narrowed down to the pattern that underlies all wars, sports com-petitions, and religious myths, namely, the ght between two sidesgoodand evilfor dominance of the world.These plots are miles away from thepsychological complexity of Victorian novels, Shakespearean dramas, andeven Hollywood thrillers, whichMurray hopes to see enacted on the Holo-deck.This predictability of the plot would constitute a weakness if narrativewere an end in itself, but it is an important asset in the case of games, sinceit allows the user to jump into the ctional world and start playing rightaway, without having to plod through tedious instructions.As was the case in Propps corpus of Russian fairy tales, individual games

dier from each other in the concrete motifs that esh out the archetypal

. This term refers to nondigital games, such asDungeons andDragons, a game that inspiredmuch of digital culture.


structure. In a predominantly visual medium, the element of narrative thatoers the richest potential for variation is the setting. This is why actiongames invest so heavily in the thrill of moving through a landscape. Butthere is another factor that accounts for the importance of spatial themes,a factor that also explains why shooting plays such an important role incomputer games. For an action game to be worth playing, the opportunitiesfor action must be frequent, or the user would become bored. As I suggestabove, living ones life is a matter of constantly engaging with the world.Moreover, players want their actions to have an immediate eect: nothingis more irritating in a game than clicking and seeing nothing happen. Butto maintain the narrative on the proper track, the range of actions must beseverely restricted. Adventure games do not preplan each possible narra-tive development, as do the Choose Your Own Adventure texts, but theymake sure that options will remain within a certain range so that the overalldestiny of the players avatar will not deviate from the general line of themaster plot. In the case of shooting, the users choices consist of selectinga weapon, aiming it, and deciding when or whether to re; in the case ofmovement, the possibilities correspond to directions, and they are limitedby the architecture of the landscape: the player can run through hallwaysbut cannot go through the walls.When players choose a direction, they seetheir avatar move immediately, and this provides the sensation of a highdegree of control. Shooting gives an even greater feeling of power becauseof the instantaneous and dramatic result of pulling the trigger. The pre-dominance of violence in computer games has been widely attributed tocultural factors, but it can be partly explained by a desire for immediateresponse.Moreover, of all human actions, none is better simulated by click-ing on a control device than pulling a trigger. It is not my intent to defendthe violence of computer games; but the theme of shooting exploits with afrightful eciency the reactive nature of the medium.

Conclusion

How, then, do I envision the future of narrative in cyberspacea phrasethat Janet Murray uses as subtitle to her Holodeck book? I will not proposea global answer to this question because digital textuality, like literature, isa eld of many genres. A reasonable coverage of the issue of digital nar-rativity should reect this diversity. I will consider three forms of digitalnarrative: the largely virtual genre of VR, or Holodeck narrative, andthe two very real genres of computer games and literary hypertext.The rst entertainment uses of VR technology will almost certainly

be action games. The playing eld will be a computer-generated three-


dimensional environment that surrounds the player, rather than a rectan-gular display restricted to the screen. VR games will implement type (internal-ontological) interactivity, and they will use a narrative frame-work as an incentive to play. But if narrative is to become the center ofinterest in a VR environment, the user should be placed in the role of theactive observer of category (internal-exploratory). This role not onlypermits a wider range of themes and emotional experiences, it is also muchmore compatible with the detachment of aesthetic contemplation than theexistential involvement of category . If digital narrative is going to becomea signicant, and reasonably popular, art form in the twenty-rst century,it will be as a movie that creates a heightened sense of presence by open-ing its world to the body of the spectator and by letting this body watch theaction from various perspectives.The two other genres, computer games and hypertext, stand at the oppo-

site ends of the cultural spectrum: one a widely popular form of entertain-ment consumed for its own sake, especially by teenage males, the other anarcane academic genre read mostly by theorists and prospective authorsby people more interested in writing about it than in reading it. (It is mainlyin this sense that hypertext turns readers into writers.) Each genre, I believe,could expand its territory by learning from the other. Though the motiva-tion of the game player is not primarily aesthetic, the care given to graphicsand to the construction of a narrative framework suggests that users arenot indierent to artistic quality. But computer games suer from the sameeconomic pressures as Hollywood movies; they are expensive to produce,and the investment can only pay o if they reach a wide audience. On theshelves of computer stores, there is only room for the gaming equivalentof John Grisham and Stephen King narratives. What is needed for com-puter games to fulll their artistic potential (and of course will not hap-pen in todays society) is an emancipation from the tyranny of the market.I can imagine games in which users would be given a concrete task butwould also be invited to take breaks in the action, during which they wouldexplore the landscape and meet characters who would entertain them withstories about the ctional world. But hardcore game players would prob-ably resent these narrative interludes as aggravating interruptions of theforwardmomentum of the game and as temporary loss of control over theirfates. The competitive involvement of the game player is basically incom-patible with the detached contemplation of the aesthetic experience, andmy proposal will only be viable if the works I am imagining are able to fos-ter a new attitude in the user, namely, the willingness to switch back andforth between the contemplative and the active stance.While the narrative variety of games has been limited by the need to


court a popular audience, the popularity of hypertext has suered from theideological contempt of its authors for the closure and coherence of classi-cal narrative. The early practitioners and theorists of hypertext thought ofthe genre as the novel of the future, but in the postmodern area, novelis more likely to mean the subversion of narrative than lengthy liter-ary narrative prose text. Since most hypertext authors aim at the highend of literary culture, they take a deliberately experimental approach tothe new writing technology.The dominant pattern of early hypertexts wasthe scrambled narrativewhat Espen Aarseth (: ) calls a game ofnarrationbut in recent years hypertext has turned toward nonnarrativetypes of meaning, such as atemporal lyrical structures, musical structures ofthemes and variations, collage of various genres, unstructured lists, visualeects, animation, and self-erasing text. It is indeed as conceptual art thathypertext has carved out for itself a modest place in contemporary literaryculture.4 The danger with the conceptual route has been clearly seen byUmberto Eco (: ): once readers have grasped the basic concept,they may feel that reading is no longer necessary. It is far from my intent todiscourage conceptual art and experimental literature, but as long as hyper-text authors limit themselves to this route, they should not be surprised tosee the medium conned to a narrow cultural niche.Hypertext cannot live forever in the academic cocoon. It will not y on its

own until it broadens its audience beyond academic circles, and it will notbroaden its audience until it learns to satisfy, rather than frustrate, narrativedesire.This does not mean that it should try to be a novel but, rather, that itshould discover narrative modes and themes more suitable to its interactivenature and multimedia capabilities. Here I must fundamentally disagreewith Robert Coover (), who thinks that the golden age of digital lit-erature came to an end when hypertext ceased to be purely verbal. To methe future of digital narrativeor more broadly, the future of digital textu-alitylies in the enhancement of verbal storytelling with visual and audiodocuments. An author who plays masterfully with the newly acquired sen-sory dimensions of digital environments is M. D. Coverley in Calia ()and The Book of Going Forth by Day, an electronic novel in progress. Thesetexts are far less fragmented than the purely textual hypertexts of the rstgeneration (e.g., Michael Joyces Afternoon), because hyperlinks can now beused to move from one medium to anothertext, pictures, musicratherthan to jump across the text.The result is a much more sustained narrative

. The Norton Antholo of Postmodern American Fiction makes room for hypertext in the post-modern literary canon by including printed excerpts from two hypertext ctions: Afternoon:A Story () by Michael Joyce and I Have Said Nothing () by J. Yellowlees Douglas.Does this amount to an ocial acceptance of hypertext, or is it a token gesture?


interest. The literary model for this new type of digital narrative is not themulticursal5 novel, such as Mark Saportas Composition No. () or Milo-rad PavicsDictionary of the Khazars (), but the artists book,6 such asTomPhillipss A Humument, or recent literary works that propose an original dia-logue between text and picture, such asThe Emigrants () by the GermanauthorW. G. Sebald. In order to take full advantage of the reactive natureof its medium, hypertext could also seek inspiration from playful art formsand artifacts that stage the act of reading as a journey lled with manysurprises: pop-up childrens books, Advent calendars, and art CDROMsthe last illustrated by the work of Norie Neumark, Agnes Hegeds, or Jean-Louis Boissier.The digital medium can give the tactile pleasure of mousingover hot spots (invisible hyperlinks) and of making images or text unex-pectedly appeara pacist alternative to the thrill of pulling the trigger inshooting games. It can tell stories in many modes and layers, by makingthe individual episodes expandable into other media or into more detailednarrations. And nally, it can handle mini-stories that ll the screen. Froma cognitive point of view, small stories are more ecient than large nar-rative patterns that need to be chunked up, because this chunking necessi-tates constant interruptions and digressions that make it very dicult forthe reader to hold onto a thread.I personally wish to see these design strategies put in the service of

projects with a do-it-yourself, cottage-industry quality that would give freerein to self-expression: projects such as building an autobiographical scrap-book, reconstructing a family saga, exploring local history, or preservingcultural memory. These projects lend themselves particularly well to thenonlinear browsing of hypertext, because the story of a life or a communityis not a dramatic narrative aimed at a climax and built on suspense butan epic narrative made of many self-sucient episodes that can be read inmany orders. Thanks to multimedia hypertext programs, such as Flash orDirector, and to the design tools of the Internet, it is now possible to tellour personal stories, or the stories of our communities, through text, music,and pictures without incurring the exorbitant costs of making a documen-tary movie or publishing a glossy illustrated book.What I am calling for isabandoning the hegemonic dream of turning newmedia narratives into the

. As Aarseth (: ) observes, a multicursal labyrinth is one that can be solved (i.e., exited)through more than one route. Similarly, a multicursal novel allows many itineraries throughthe text.. Artists books are visual artworks that present themselves as a collection of illustrated pagesbound together in book form. Some are meant to be printed, while others remain originalmanuscripts. In either case, the work reveals itself to the spectator through the activity ofpage turning, and this dynamic mode of apprehension becomes a source of artistic eects.See on the topic Drucker and Hubert and Hubert .


art (read: the highbrow, avant-garde art) or into the entertainment form (read:the mass entertainment form) of this new century and seeking for these narra-tives a less glamorous, but no less important, place in culturea place thatwill represent a true democratization of digital textuality.

The claims made in this essay can be summarized by three seeminglyincompatible statements:

. The truly distinctive feature of digital media is interactivity.This fea-ture enables the user to choose her or his way through the text atrun time.

. Interactivity does not make it easy to tell stories, because a narrativeinterpretation is a response to a linear structure that is built into thetext, not a type of meaning freely created by the reader out of any setof data.

. Yet without some degree of narrativity, digital media cannot becomea major presence on the arts and entertainment scene.

Digitality is a uid environment; narrative, as a type of meaning, is a solidstructure. To reconcile the two, some compromise will be necessary. Nar-rative will have to learn to share the spotlight with other types of sensorydata; to accept a subordinate role, as in games, or limit itself to certain plottypes. Conversely, the medium will have to give up some of its uidity toallow narrative meaning to solidify in the mind of the reader. This meanslimiting the range of possible actions, channeling interactivity, and neu-tralizing the threat that it poses to coherence by orchestrating periods ofuser activity and periods of system control. If digital narrative is goingto be more than a new mode of diusion for texts that could be material-ized in print, such as the works of Stephen King, it cannot be a freewaythat takes the reader through the landscape, as do standard novels. But ifit is to keep narrative desire alive, it cannot be a wilderness, where linksare so numerous that the reader is lost in a thicket that looks the same fromevery position. To borrow a metaphor from Mark Bernstein of EastgateSystems (), the compromise between being lost in the wilderness andbeing sucked onto the freeway is to be invited into a gardenwithmany care-fully designed paths. These paths guide users through the narrative land-scape and enable them to see it from various points of view without losingtheir sense of orientation. But rather than making the experience fully pre-dictable, the paths reveal unexpected, delightful features at every turngazebos, follies, grottoes, statues. This combination of designed space andserendipitous discovery, mapped trails and surprise attractions, containedarea and expanding vistas make the garden look much bigger than it reallyis.This may be the closest one gets to the mythical Aleph, without entering


a jungle where narrative meaning chokes in the brambles of uncontrollablemultiplicity.

References

Aarseth, Espen Cybertext: Perspectives on Ergodic Literature (Baltimore, MD: Johns Hopkins Univer-sity Press).

Amerika, Mark Grammatron.Web site accessed at www.grammatron.com, September .

Auletta, Ken What I Did at Summer Camp, New Yorker, July , .

Bernstein, Mark Hypertext Gardens: Delightful Vistas. Eastgate Systems Web site, accessed atwww.eastgate.com/garden, February .

Bolter, Jay DavidWriting Space: The Computer, Hypertext, and theHistory of Writing (Hillsdale,NJ: LawrenceErlbaum).

Borges, Jorge Luis The Aleph and Other Stories, edited and translated by Norman Thomas di Giovanni(New York: E. P. Dutton).

Campbell, Joseph [] The Hero with a Thousand Faces, d ed. (Princeton, NJ: Princeton UniversityPress).

Cassell, Justine, and Henry Jenkins, eds. From Barbie to Mortal Kombat (Cambridge: MIT Press).

Coover, Robert Literary Hypertext: The Passing of the Golden Age, Feed Magazine, February .Web magazine accessed at www.feedmag.com/document/dolo.html, March .

Coverley, M. D. Calia (Watertown, MA: Eastgate Systems).in progressTheBook of Going Forth byDay. Accessed at calia.hispeed.com/Egypt, August ,.

Ditlea, Steve False Starts Aside,Virtual Reality Finds New Roles, New York mes, March , C.

Don, Abbe Narrative and the Interface, in The Art of Human-Computer Interface Design, edited byBrenda Laurel, (Redding, MA: AddisonWesley).

Douglas, J. Yellowlees I Have Said Nothing. Hypertext software. Eastgate Quarterly Review of Hypertext. (Cambridge, MA: Eastgate Systems).

Drucker, Johanna The Century of Artists Books (New York: Granary).

Eco,Umberto The Open Work (Cambridge: Harvard University Press).

Ferguson, Niall, ed. Virtual History: Alternatives and Counterfactuals (London: Macmillan).

Geyh, Paula, Fred G. Leebron, and Andrew Levy, eds. Postmodern American Fiction: A Norton Antholo (New York: W.W. Norton).

Heim, Michael The Metaphysics of Virtual Reality (New York: Oxford University Press).


Hubert, Rene Riese, and Judd Hubert The Cutting Edge of Reading: Artists Books (New York: Granary).

Iser,Wolfgang The Reading Process, in Reader-Response Criticism: From Formalism to Poststructuralism,edited by Jane P. Tomkins, (Baltimore, MD: Johns Hopkins University Press).

Jakobson, Roman Linguistics and Poetics, in Style and Language, edited by Thomas Sebeok, (Cambridge: MIT Press).

Johnson, Steven Interface Culture: How New Technolo Transforms the Way We Create and Communicate (SanFrancisco: HarperEdge).

Joyce, Michael Afternoon: A Story. Hypertext software (Watertown, MA: Eastgate Systems). Of Two Minds: Hypertext, Pedago, and Poetics (Ann Arbor: University of MichiganPress).

Landow, George P. Hypertext .: The Convergence of Contemporary Critical Theory and Technolo (Baltimore,MD: Johns Hopkins University Press).

Larsen, Deena Marble Springs. Hypertext software (Cambridge, MA: Eastgate Systems).

Laurel, Brenda Computers as Theatre (Menlo Park, CA: AddisonWesley). Art and Activism in VR,Wide Angle (): .

Laurel, Brenda, ed. The Art of Human-Computer Interface Design (Redding, MA: AddisonWesley).

McHale, Brian Postmodernist Fiction (New York: Methuen).

Moulthrop, Stuart Victory Garden. Hypertext software (Watertown, MA: Eastgate Systems).

Murray, Janet H. Hamlet on the Holodeck: The Future of Narrative in Cyberspace (New York: Free Press).

Olafson, Peter Game Theory, New York mes, October .

Pavel, Thomas Fictional Worlds (Cambridge: Harvard University Press).

Pavic, Milorad Dictionary of the Khazars: A Lexicon Novel in , Words, translated by ChristinaPribicevic-Zoric (New York: Knopf ).

Phillips, Tom A Humument: A Treated Victorian Novel (London: Thames and Hudson).

Propp,Vladimir [] Morpholo of the Folktale, translated by Laurence Scott (Austin: University ofTexas Press).

Ryan, Marie-Laure Narrative as Virtual Reality: Immersion and Interactivity in Literature and Electronic Media(Baltimore, MD: Johns Hopkins University Press).

Saporta, Marc Composition No. (Paris: Seuil).

Sebald,W. G. The Emigrants, translated by Michael Hulse (New York: New Directions).

Walton, Kendall Mimesis as Make-Believe: On the Foundations of the Representational Arts (Cambridge: Har-vard University Press).

beyond myth and metaphor- narrative in digital media

Documents

digital narrative

literary narrative

concept of narrative

narrative structures

universal narrative

digital medium

digital mediaryan

digital narrativity