Image and Narrative
Online Magazine of the Visual Narrative - ISSN 1780-678X



Home archive




Issue 12: Opening Peter Greenaway's Tulse Luper Suitcases

Virtual Recentering: Computer Games and Possible Worlds Theory

Author: Jan Van Looy
Published: August 2005

Abstract (E): This article is an attempt to evaluate the benefits of using a possible worlds framework to describe virtual worlds as they occur in computer simulations and games. Although the use of fiction theory to describe digital media is not always without its problems, it can function as a starting point to analyze the specificity of new forms like computer games. Possible worlds logic is a theoretical framework introduced in the 1960s by Saul Kripke in order to be able to assess the truth-value in counterfactual statements (if... then...) and has later been applied to fiction by theorists like Thomas Pavel and Marie-Laure Ryan. Fictional recentering is the move by which a reader is invited to step into a possible world where a substitute speaker narrates events that have taken or are taking place in the fictional world. When analyzing digital media, it is possible to observe a similar process, which could be called virtual recentering, and which is both similar to and different from fictional recentering. Finally, by way of illustration, the computer game Myst will be analyzed using the framework sketched earlier and will be interpreted as a comment upon its own format and its inherent independence from its makers.

Abstract (F): Cet article se propose d'évaluer les avantages de la théorie des mondes possibles pour la description des mondes virtuels tels qu'on les trouve dans les simulations informatisées ou les jeux vidéo. Bien que le recours à une théorie de la fiction pose des problèmes quand on la transpose aux médias numériques, elle peut fournir un bon point de départ à l'analyse des traits spécifiques de nouvelles formes culturelles comme les jeux vidéo. La logique des mondes possibles est une théorie qui a été introduite dans les années 60 par Saul Kripke afin de résoudre les problèmes posés par la véridicité de certains énoncés non factuels (si. alors.) et qui a été appliqué aux textes de fiction par des auteurs tels que Thomas Pavel et Marie-Laure Ryan. On appelle « recentrage fictionnel » le geste par lequel un lecteur est invité à entrer dans un monde possible où un narrateur autonome raconte des évéments qui se passent ou se sont passés dans cet univers de fiction. L'analyse des médias numériques révèle un phénomène analogue, c'est-à-dire comparable mais non identique, que nous proposons de nommer « recentrage virtuel ». Afin d'illustrer pratiquement les idées théoriques exposées dans la première partie de l'article, notre texte se termine par une analyse du jeu vidéo Myst . Plus concrètement, nous analyserons comment ce jeu fournit aussi un commentaire sur sa propre forme et l'indépendance qu'il acquiert par rapport à ses producteurs.

keywords: computer games, possible world semantics, narratology, fiction theory, game studies


When a new cultural artifact is introduced into academic research, existing disciplines tend to behave imperialistically. When new media art and digital culture arrived, this is precisely what happened. Disciplines as diverse as literary theory, cultural studies, psychology, sociology, art history, design, engineering, and computer science attempted to cover as much terrain as possible. The resulting method wars created misunderstanding, but also a number of interesting discussions. It is to one of these discussions I would like to turn in this article: that of the usefulness of literary theory - narratology more precisely - for computer game research. I find new media theory too vast a field to cover and the essentialist idea of 'the computer medium' as one singular structure of well-defined characteristics untenable. However, I believe that computer game research is a fast-growing and significant project both for game researchers and the humanities at large. In view of the scope of this issue I will take the perspective of the literary scholar (which, partly, I am still) who is interested in the workings and specificities of narrative in computer games.

Probably the first - and arguably the most important - literary scholar who has developed an interest in computer games is Espen Aarseth. In Cybertext: Perspectives on Ergodic Literature (1997), he devises a theory of play primarily focusing on texts (as opposed to graphical media). While he draws extensively on literary theory from Jakobson to Iser, he is careful not to lose track of his research object. He explicitly challenges the application of theories of literary criticism to new empirical fields when this is done without reassessment of the terms and concepts involved (14). In this way, he implicitly refers to the traps hypertext theorists have set for themselves from the late 1980s onward. This reluctant attitude towards narrative theory for games is not uncommon in computer games research today. Jesper Juul (2001), for example, not only criticizes the manner in which some literary concepts have been inappropriately applied to computer games, but also the fact that they have been applied at all. He calls this process biased because "it emphasizes some traits and suppresses others" and "may bear hidden assumptions." Particularly, he laments the description of games as narratives when this is done only because narrative is a cognitive mechanism which we use to make sense of the world and because it is fundamental to human thought. The fact that we speak about a game in the form of a story is not enough for it to be one. He calls this a "holistic view" and an "a priori argument" (Juul 2001).

While I do agree with much criticism regarding the way in which literary theory has been used and abused to describe the functioning of new media, I do have problems with the general tenor of the argument. The reason for this is the way in which the problematics is approached. I believe that the question in point should not be whether a computer game is a narrative or not, but rather to what extent it bears its characteristics. Like in film and literature, in games there are many more aspects involved than just the story structure. With Marie-Laure Ryan, I believe that the inability of narratology to account for the full experience of games does not mean that we should do away with the concept of narrative in ludology altogether. Many theorists argue that computer games are played for the sake of solving problems, defeating opponents or developing certain skills and not for the purpose of creating a 'trace' that reads as a story. However, if narrative were entirely irrelevant to games, why would designers put so much effort in the narrative framework? Ryan's explanation is simple, but adequate: narrativity is not the 'raison d'être' of games, but it plays an important role as a stimulant for the imagination. Film clips and story material are used to immerse the player into the virtual environment. It seems that the concept of 'narrative' like that of 'interactivity' is unable to describe the phenomenon of game play entirely, but it should be taken into account when developing a theory of gaming.

Ryan - on whose work I will be drawing extensively in this study - makes a difference between fiction and narrative theory. She explains it as follows: "While fiction is a mode of travel into textual space, narrative is a travel within the confines of this space" (Ryan 1991: 5). Fiction theory describes the relationship between the actual world and the world evoked by the text. Which intermediaries exist between the reader and the events described? Why do we regard certain formats as fiction? What does it mean for a text to be fictional and how does it relate to the world we live in? Narrative theory , on the other hand describes the relations between the components within the narrative world. Which is the chronology of events? How is the tension built up and maintained? What is the role of the different characters? Their wishes? Their prejudices? In this article, I will be dealing with fiction rather than narrative. Starting from the theory of possible worlds I will investigate how digital media and especially computer games make use of the human capacity to imagine possible environments. After a brief history and description of possible worlds theory, I will look at certain aspects in fiction (I will reserve the term 'fiction' strictly for textual fiction, as opposed to 'virtual' computer games) that have been described using its logical framework, and I will compare this to the way in which digital media deal with possibility. Finally, I will propose a framework for virtual recentering and illustrate this by looking at the computer adventure game Myst .



Possible Worlds


The theory of possible worlds is a formal model developed by logicians for the purpose of describing the semantics of modal operators - primarily necessity and possibility. In traditional truth-functional logic, it is impossible to discuss the truth-value of counterfactual statements like "If I had been an Inca, I would have invented the wheel" in a satisfying manner. The answer would simply be that you are not an Inca and therefore anything can be said ('ex falso sequitur quodlibet', from a false proposition anything follows). This is counterintuitive since the speaker of the sentence may want to make a point about himself. In modal logic, however, the propositional constituent "If I had been an Inca" creates a passage to a possible non-actual world in which the I is an Inca and automatically situates the second part of the claim in this construct. "In a certain possible world I am an Inca and I have invented the wheel." In this way it is still impossible to objectively determine whether the 'I' would have done what she claims, but an acceptable framework for discussion is created.

This has inspired literary theorists to apply a similar frame to fiction. While there is no explicit world-creating operator in fiction, the knowledge of the fictional status of the text suffices to create the passage sketched above. There is no explicit introduction like "Dear reader, this is a novel, the events are therefore fictional." Nevertheless, the reader is implicitly urged to adopt a certain style of reading and interpretation. A possible world is presented in which certain events take place which can be discussed and interpreted. "The theory [of possible worlds] has two concepts to propose to textual semiotics: the metaphor of 'world' to describe the semantic domain projected by the text; and the concept of modality to describe and classify the various ways of existing of the objects, states, and events that make up the semantic domain" (Ryan 1991: 3).

One of the first philosophers to approach fiction as a logical issue was Gottlob Frege (1848-1925). According to Frege, a sentence about an imaginary object, place or time does not automatically entail that the statement is false. Rather, it should be excluded from the set of true statements on grounds of referential failure (while errors and lies illustrate the case of faulty predication). This approach is problematic, however, when applied to sentences referring both to actual and fictional entities, for example "Conan Doyle created Sherlock Holmes." According to Frege's theory, Sherlock Holmes cannot be referred to and therefore the above statement cannot be true (which of course it is). To avoid this, we must separate reference from existence and accept that we can refer to objects that do not exist in the actual world. In order to frame this type of reference, twentieth-century philosophers returned to the Leibnizian concept of possible worlds. According to Leibniz, an infinity of possible worlds exist as thoughts in the mind of God and only one is actual: the best of them all, chosen by the divine mind to be instantiated (Ryan 1991: 18). Language is able to refer to these possible worlds and fiction is one means to do so.

But what is the ontological status of a possible world in an utterance? Ryan distinguishes between two major positions. On the one hand there is David Lewis who professes modal realism . For Lewis, possible worlds are what they are and there is no ontological difference between them. The actual world is just the realization of one possible world among others. While to us one world may appear more acceptable than another because it is closer to the world we live in, according to Lewis there is no intrinsic difference and we should treat all possible worlds as equally probable. On the other hand, there is the modal fictionalism or anti-realism of Gideon Rosen and Nicholas Rescher. The latter equates actuality with reality, restoring the uniqueness of the actual world. Possible worlds are constructs of the mind rather than absolutely existing entities. He does not start from the collection of possible worlds to then say that the actual world is just the realization of one of them. Rather, he starts from the actual world and claims that possible worlds are abstract properties to make sense of and analyze the world we live in. Ryan finds what she calls a 'compromise' between the realist and the fictionalist position in Kendall Walton's comparison of fiction to a game of make-believe. Children engaging in make-believe agree on a number of rules of substitution (e.g. a bucket of sand becomes a pie) through a semantic operator indicating pretense: "Let's pretend that the bucket is a pie." While in make-believe there is often an explicit operator like "Let's play family" opening the passage to a possible world, in fiction this operator is more often implicit than not. The reader accepts the textual universe as an imaginary alternative to her system of reality, and for the duration of the 'game' of reading she steps into the world created by the book and accepts the proposed substitutions. "Through its double perspective on the textual universe, the make-believe approach to fiction reconciles our intuitive belief in the unique character of the actual world and in the privileged character of our point of view with our willingness to suspend the disbeliefs entailed by this belief (Ryan 1991: 23).



Virtual recentering


When a reader follows the passage into a fictional world, the realm of possibilities is recentered around her. For the duration of her immersion, she accepts the sphere created by the narrator as the actual world in the same way that the interlocutor will try to imagine the other as an Inca and the children will take the bucket for a pie. For Ryan, this fictional recentering presupposes three modal systems and three actual worlds. The first is our native system with the actual world (actually actual world) at its center. Textual fiction provides a passage to a second universe with the world projected by the text at its center (textual actual world). The third, then, is the system to which the text refers , a system that contains everything projected by the text, but also everything that is not mentioned and is filled in by the reader (textual reference world). While in fiction the world described by the text is always different from the actual world (as opposed to non-fiction), it will generally be indistinguishable from the world to which it refers (the third system). Only when a narrator lies and this can be inferred, the reader will know that what is described by the text is not what the world it refers to is like. Since this is a marginal case and insignificant for this study, I will focus on two systems when dealing with recentering: the actual world on the one hand, and the textual world on the other. Moreover, I tend to take the poststructuralist stance that the actual world is equally a cultural construct. It should be seen as a representation conforming to the reader's image of the world, not as something that exists anywhere outside of representation. This perspective permits to install a symmetry between two representations: the representation of the actual world present in the head of the reader, and the representation of the textual world built from the text, and it explains why the reader can slip from one world into another so seamlessly.

At this point I would like to move from textual fiction to digital media. Both logicians and narratologists seem to regard (fictional) recentering as a matter of language, but I would like to contend this view. When playing a computer game, or interacting with an installation for that matter, immersion is imminent. In the process of virtual recentering , the player (I will refer to all instances of user, interactor, wreader etc. as 'player') accepts a movement into a virtual environment in very much the same way a reader glides into the fictional world. This seems to indicate that recentering is not so much a matter of language, but of structures more basically cognitive, a semantic layer neither independent from, nor entirely determined by language. The player willingly suspends disbelief and is ported to a world created in the virtual. Her cognitive system accepts the fact that there is a possible world where she is a ninja, a god, a railroad tycoon or a spaceship. And what is more, she accepts that she has only very limited control over this persona. She accepts certain key combinations to be actions like children accept buckets to be pies. She accepts changing pixels to be dungeons, monsters or whole armies just like a reader accepts sequences of words to be landscapes, grails, Sancho Panza... In this respect Aarseth speaks of functional 'autism'. "The contract between user and text in 'interactive fiction' is not merely a 'willing suspension of disbelief', but a willing suspension of one's normal capacity for language, physical aptness, and social interaction" (Aarseth 1997: 116-117).

This does not mean, however, that virtual recentering is identical to fictional recentering. The most important difference between the two is that whereas fictional recentering consistently opens the gate to the same world, virtual recentering does not. It is true that texts can be read from different perspectives entailing different analyses, but the artifact retains its shape. Letters, words, sentences remain at their respective positions. The material text remains the same and so do the possible worlds it generates. Each virtual recentering, however, creates a single and unique possible world. Even in very simple games, a player will hardly ever be confronted with the same situation twice due to the easiness with which a computer can take into account high numbers of parameters. Each event takes place at a unique time, both in the actual and the virtual world. Events in a work of fiction happen only once, even when you reread the book. This very basic fact constitutes one of the most important aspects of the pleasure derived from playing a computer game. It is precisely the trying out of the myriad possible variations and the reaction of the environment that lies at the center of game entertainment. I do not believe that one type of recentering is more basic than the other (fictional or virtual). I would like to postulate a cognitive mechanism of recentering that forms the basis of and allows for both fictional and virtual recentering. Finally, I would like to refer to Lev Manovich who identifies a phenomenon highly similar to virtual recentering as the core aesthetic of digital artifacts: "a constant, repetitive oscillation between an illusion and its suspense." (Manovich 2001: 205; see also Van Looy, 2003).



Minimal departure


Relations between worlds are always modal relations. A textual or virtual world is only acceptable as such when it is 'possible' (hence 'possible worlds'), when it satisfies the logical laws of noncontradiction and of excluded middle. The law of noncontradiction states that something cannot simultaneously be true and false (for proposition P, it is impossible for both P and 'not P' to be true). The law of excluded middle states that something is either true or false (either P is true or 'not P', one of both). E.g. a world is possible when it either contains human characters or it does not (not both at the same time; law of noncontradiction), and it has to be one of both (law of excluded middle). But there is more at stake. According to Saul A. Kripke, possibility is synonymous with accessibility : a 'destination world' is possible in a system of reality if it is accessible from the 'source world'; logical compatibility is just one aspect of this relation. Ryan identifies some eight other accessibility relations such as 'identity of properties' (objects in the destination world have the same properties as those in the source world), 'identity of inventory' (both worlds are furnished by the same objects) and physical compatibility (both worlds share natural laws). Thus she molds the rigid and simplifying logical model into a powerful framework allowing for thorough analyses of fictional and virtual constellations (31).

But how can a fictional world be 'possible'? How can the law of excluded middle apply to its entirety, to all it contains? How do we know, for example, that there are fleas in the world depicted by Gustave Flaubert in Madame Bovary (1856) when they are not mentioned at any point? Ryan's response to this problem is that there must be a mechanism allowing the reader to combine information from the text with other sources, primarily the actual world. The reader reconstrues the textual world in the same way children construe their game world: as conforming as far as possible to the actual world. Thus a bucket of sand turns into a pie, but the dog remains the dog; unicorns have a horn, but they also have four legs like ordinary horses. This is what Ryan calls the principle of minimal departure : Readers project upon the textual world everything they know about reality, and make only the adjustments dictated by the text. "It is by virtue of the principle of minimal departure that readers are able to form reasonably comprehensive representations of the foreign worlds created through discourse, even though the verbal representation of these worlds is always incomplete" (52).

Interestingly, the principle of minimal departure applies to virtual recentering in a similar fashion. When the player arrives in a virtual world, she projects her knowledge onto her surroundings. It is only gradually that she will discover those elements that do not conform. And even then, when the player finds a telephone that she cannot operate, for example, she will notice that it is neither part of the simulation nor the game. Nevertheless, on a subconscious level she will attribute it to the virtual world. After all, it is not because the player cannot operate the telephone, that the characters or the inhabitants of the virtual world do not. As a third factor in the reader's creation of an image of a fictional world (beside the text and the actual world), Ryan mentions intertextuality and generic landscapes. Readers not only project properties from the actual world, but also from possible worlds depicted in other texts, especially those belonging to the same genre. Again, the same applies to virtual worlds which will always also be treated according to earlier experiences with other virtual worlds.

One group of elements in the (textual) fictional world that is not susceptible to the principle of minimal departure is that of indexical clues, more particularly first- and second-person pronouns. Only a naïve reader will directly identify the 'I' as the author, and hardly any reader at all will identify a 'you' as himself. As will be further examined in the next section, the author has a representative in the fictional world in the form of a narrator or substitute speaker which is referred to by the pronoun 'I' and the reader is represented by the substitute hearer, the instance that receives the fictional communication and to which is referred by the pronoun 'you'. For Ryan, this referential divorce between the I of the actual world and the relocated I of the textual world is the distinguishing characteristic of fictional communication. However, when we look at virtual recentering, we see that the first person pronoun is susceptible to the principle of minimal departure. Thus, a player will always refer to his persona or avatar as 'I' ("Then I went into the cave") whereas a reader of a work of fiction will never refer to the main protagonist as 'I' - even when the novel is written in the I-form. In the previous section I have identified the unique temporality, the fact that each passage into the virtual creates a unique possible world, as central to digital aesthetics. The fact that the 'I' is consistently submitted to the principle of minimal departure is another, perhaps an even more fundamental characteristic of virtual recentering.



Intrigue - intrigant - intriguee


Typically, each instance of fictional communication involves a recentering to a possible world. Sender and receiver accept to take part in a game of pretense: "Let's pretend that this and that actually happened to him and her..." Without explicit predication like in counterfactual statements, but with knowledge of fictional intention the reader follows the passage to where fictional events take place. But how does this dual-world, dual-intent, single-text structure function? How does the author's implicit utterance embed, frame and relay the discourse attributed to the speaker in the fictional world? Ryan describes this problem in intentional terms using John Searle's illocutionary framework (1975). Starting from fiction as a speech act of assertion she identifies an actual speaker (author) and an actual hearer (reader) who decide to pretend that they are two people (substitute speaker and substitute hearer) having a real communication about real facts (in the textual world). The author pretends to be a substitute speaker and the reader is invited to pretend to be a member of the world where this substitute speaker is communicating with a substitute hearer. If we formulate this constellation starting from the intention of the author we see that she is an actual speaker who, by producing the text, invites an actual hearer (reader) to pretend that there is a substitute speaker transmitting a real text about real events to the substitute hearer in the textual world (66-75).

In fiction, there are two varieties of substitute speakers. Personal or individuated narrators have a psychological reality and may assume a prominent role in the process of narration and in the narrative itself. The history of Machu Picchu could be told using the 'I'-perspective for example. The substitute speaker will then refer to himself as 'I' both in the process of narration ("I will now tell you the story of Pachacutti") and in the ongoing action ("I watched Pachacutti ride into Machu Picchu"). Impersonal narrators on the other hand do not have a psychological reality; they remain out of the picture both in action and narration ("A large crowd watched Pachacutti ride into Machu Picchu.") and their existence is postulated on purely logical grounds. In order to be able to retain the fictional layout sketched above, Ryan inserts a dummy into the slot of the substitute speaker when the text prevents the individuation of the narrator, when there is no identifiable instance mediating the events (70). The function of this dummy substitute is to prevent coreference between the 'I' of the actual world (author) and the 'I' of the textual world (substitute speaker), protecting the impersonal narrator from the principle of minimal departure (which is required in fiction). While a personal narrator may be unreliable, an impersonal narrator may not, because the gap between the truth and the narrator's declarations could not be justified on psychological grounds. Finally, on the side of the substitute hearer we have a similar configuration. The written communication can be addressed to a specific instance 'you', e.g. in epistolary novels, but most of the time the substitute hearer slot is indeterminate and should be filled with a dummy receiver.

When we look at virtual recentering, we see that there are several parallels. Like in fictional recentering, the actual receiver (player) is expected to follow the passage into a possible world. She is invited to engage in a game of pretense ("Let's pretend that you are in a place that looks so and so and where you have to do this and that"). There is a representative of the player in the virtual world: the substitute receiver. It becomes more problematic, however, when we look at the substitute speaker. In most computer games and virtual environments there is no trace of such an instance. Moreover, intuitively it is very difficult to think about playing a game just in terms of communication. In fiction, the author creates a voice in the fictional world which pretends that what it is saying is true. In a computer game there can be a guiding voice, but this voice is not the omnipresent ruler of the virtual world. It is nothing more than a guide while the virtual world is experienced more directly, in most cases without the intervention of language. The emphasis in the virtual world shifts from substitute sender (fiction) to substitute receiver (computer game). It is the latter that primarily functions as an information filter for the player. But how do we explain the feeling that we sense a creator while playing a computer game? There must be some way of explaining the sense of not being alone in the virtual world. Sometimes this other instance is no more than the virtual world itself, in other cases there are jokes and wisecracks deliberately aimed at the speaker. How do we account for them? How do we explain the workings of the virtual world and its substitute sender and receiver?

When discussing adventure games Espen Aarseth introduces the technical concept of intrigue . He argues that the adventure game effectively disintegrates or at least weakens the notion of story by forcing the player's attention on the elusive plot. "Instead of a narrated plot, cybertext produces a sequence of oscillating activities effectuated (but certainly not controlled) by the user. But there is nevertheless a structuring element in these texts which in some way does the controlling or at least motivates it. As a new term for this element I propose intrigue, to suggest a secret plot in which the user is the innocent, but voluntary, target (.) with an outcome that is not yet decided" (Aarseth 1997: 112). As explained above, in a virtual environment the emphasis shifts to the receiver. Aarseth commences his discussion by identifying the target of the intrigue, the substitute receiver, as the intriguee . This intriguee is a parallel to the narratee of narratology, but she has a different role. While the substitute hearer slot in fiction is often filled by a dummy and is therefore insignificant to the narrative itself, in a computer game the intriguee is an integral part of the work. Without an active intriguee there is no play. The position of the intriguee is transcendental. It depends on the identification, the cognitive merger between player and puppet / avatar. When the game is over, the union dissolves (113).

As a corollary to the intriguee, Aarseth proposes the intrigant , the substitute sender in the computer game. The intrigant is the architect of the intrigue, "the mastermind ultimately responsible for events and existents, but who is not motivated by any particular outcome." This intrigant is not some sort of narrating instance. There is no character or well-defined entity that assumes its role. Rather, it is comparable to an implied author in fiction; it is more a negotiator, an abstract intermediary than a concrete voice, a mechanical construct in a real sense (114). If we take Pac-Man (Burden 1983) for example, the intriguee is the player controlling the dot-eating puppet. The intrigant is the equally transcendental entity that controls the ghosts, increases the speed and determines that when a ghost catches the player, she loses a life. For an experienced player, the intrigant is the true opponent in the game. She will peel off the narrative layer and attempt to understand according to which parameters the ghosts are moving. In the beginning, the player will be guided primarily by the story and the indexical clues. After a while, she will attempt to look beyond the surface into the ergodic structure itself. She will try to control the intrigant and anticipate its actions.

The distance between actual and substitute receiver is variable. In some games like Mario Bros (Shortt & Lowe 1986), the avatar has a name and specific characteristics. In other games such as Doom (Romero & Carmack 1993) or Flight Simulator (Artwick 1980) there is no significant intermediary character. In puzzle games like Tetris (Jones-Steele 1987) it can even be argued that there is no distance at all and therefore no virtual recentering, no creation of a possible world. I believe there is since the player is invited to believe that piling blocks and setting high scores can be important in some way. Like in Ryan's fictional framework, I prefer to fill all slots of the virtual configuration. Even when there is no visual puppet (like in Tetris), I prefer to insert a dummy intriguee. This is one explanation of why computer games can seem so immoral. It is not the player who kicks, shoots, kills, wages war, and causes environmental disaster. It is a transcendental entity 'player / puppet' who deliberately pushes the intrigue to its limits in order to learn the ways of the virtual world (see also Newman 2002). When, as a game designer, you try to build a reactive environment you automatically arrive at conflict as its major driving force. I do not believe that computer gamers are intrinsically more violent than fiction readers. It is the interaction of the format that calls for simple implementations of the conflict principle while in fiction this can be done in a more sophisticated manner. Perhaps future developments in computer (game) design will allow the genre to move in this direction.





In order to illustrate the functioning of virtual recentering in general and the concept of intrigue in particular, we shall now turn to the adventure game Myst (Miller 1994). Adventure games can be divided in two large groups based on their interface: text adventures and graphical adventures. Probably the first computer text adventure ever was Advent(ure) created by William Crowther and further developed by Don Woods (1976). It was based on Gary Gygax's popular role-playing board game, Dungeons and Dragons (1974). In the late 1970s a group of programmers from MIT's Artificial Intelligence Lab formed the company Infocom which was highly successful at producing and marketing the text adventure genre, e.g. with Zork created by Marc Blank and Dave Lebling (1977) which is said to have sold a million copies (Aarseth 1997: 101). In the early 1980s computer graphics became better and cheaper and the adventure game genre, with its spatially oriented themes of travel and discovery gradually migrated from text to images and later to three-dimensional realtime generated graphics, e.g. Syberia (Sokal 2002). The basic principles remain the same however. Marek Bronstring defines the genre of 'adventure game' as typically focusing on puzzle-solving and avoiding action elements (2002). The adventure game is the most narrative-driven computer game genre.

Myst was created by Rand Miller and his brother Robyn under the corporate name of Cyan, an independent software company they founded to develop games and educational software under the aegis of Broderbund Software, now owned by Ubi Soft Entertainment. Myst is said to have sold more than 1.5 million copies and is generally regarded as perhaps "not the greatest adventure of all time, but (...) the first great one."[1] It is a first-person (the player sees through the eyes of the puppet) exploration adventure with an emphasis on the graphically immersive environments and the 'brain-teasing' puzzles as opposed to third-person adventures such as Grim Fandango (Schafer 1998) and King's Quest (Williams 1984) where the emphasis is more on character development and interaction (Bronstring 2002). It has a classical star-shaped plot-line (one central world and a number of satellites) and it is more episodic than dramatic in that the puzzles have to be dealt with in order, not in parallel or in a sequence chosen by the reader. Somewhat like in the picaresque the episodes have little or no relevance to each other, except for the fact that they are part of the development of the main protagonist (see also Aarseth 1997: 124-125). I will be dealing with Myst: Masterpiece Edition - which is a graphically enhanced version released in 2001 - as it is included in the Myst Trilogy collection box.

The theme of Myst is already introduced when you look at the cardboard Trilogy box, which is shaped like a book with a cover, a binding and the fake imprint of pages at the side. When you start the program there is no endless beginning sequence explaining what the game is all about, only a short animation showing a human figure falling into an abyss, vanishing and leaving a book tumbling through space. An offscreen voice speaks about the "Myst book" and whose hands it might fall into. The volume lands amid darkness and the game starts. When he was young, Atrus was taken to the land of the D'ni where his father taught him how to create worlds in books and then bring them to life. One of these books is the 'Myst book' with a subtle allusion to the creation of the game by the designers. When the game starts, you receive a message from Atrus explaining how the books he created not only described fantastic worlds, but also offered the reader the chance to step into them. Something has gone terribly wrong however. One by one, his worlds are being destroyed by someone's greed, probably one of his sons: Sirrus or Achenar.[2] The player's task is to enter these worlds and restore them in their initial shape. This is done by walking around the different islands, solving the puzzles and finding the missing pages of the books in the Myst library. To move from one island to the other you have to find the so called 'linking books' which, after an often gorgeous overview of the destination world from a helicopter perspective, transport the player to its beginning.

The reason for my adding Myst as a case-study and illustration is that it can be interpreted as a comment on its own format and on the virtual structure of a game in general. By constantly referring to the book as a material object and to its fictional worlds, Myst demonstrates a preoccupation with the concepts of 'medium' and 'virtuality'. Atrus is put forward as the author of the worlds which the player is to repair. In a sense, he is the representative of the creator of the game, just like the narrator would be the mouthpiece of the author in a book. In fiction, all events are mediated through the substitute speaker. This condition appears to be artificially recreated here. The worlds exist only in the magical books created according to the ancient rites of the D'Ni. However, despite the imitation of the fictional layout, the narrative is not the most important element in the game. Rather, it is part of a larger intrigue consisting of story, exploration and puzzles, which together constitute the structure and produce the game experience. As explained before, Myst is a first-person adventure and the presence of the puppet is reduced to a minimum. Nevertheless, the player assumes a role in the virtual world which is more than his actual self. The notes from Atrus, for example, are addressed to a person whom he knows (which is you). This makes you feel that you are not just yourself visiting Myst , you unify with the puppet and take the part of the intriguee. Finally, despite the attempts to fit in Atrus as the mastermind of the game, the player does not have the impression that it is Atrus who is triggering the events. Rather, the transcendental instance of the intrigant is presenting the world to the intriguee. Myst is a game about a creator who has lost control over his worlds, just like game designers have today lost control over their creations.





Using narratological concepts and theory to describe digital media is not always unproblematic, but it can deliver valuable insights both into the nature of fiction theory and of digital media. Possible worlds theory is a logical framework primarily devised to evaluate the truth-value in counterfactual statements, and has consequentially been used to describe the functioning of fictional recentering. A reader is invited to step into a possible world where the author, or rather her substitute speaker, recounts certain events that have taken or are taking place in the textual world. When we look at digital media, we can discern a similar process: virtual recentering. However, while fictional recentering always opens a gate to the same possible world, virtual recentering each time creates a single and unique instance. Moreover, unlike in fiction, the personal instance 'I' is susceptible to the principle of minimal departure, i.e. the fact that a reader or player always relies on her knowledge of the actual world to approach the virtual. While a computer game player will refer to her avatar or intriguee as I, the fiction reader will not generally regard the main protagonist as herself. When we look at the substitute sender in virtual 'communication', we see that she is not a narratorial instance, but rather a transcendental entity pulling the game or intrigue's strings. There is no voice recounting the events, but a number-crunching intrigant generating them. Finally, when we look at the computer adventure game Myst , we see that it can be interpreted as a comment on its own format. The narrator no longer exists and the author loses control over her creation.



Cited Works


Aarseth, Espen J. (1997). Cybertext: Perspectives on Ergodic Literature . Baltimore, London: The Johns Hopkins University Press.

Artwick, Bruce (1980). Flight Simulator . Computer game. subLOGIC.

Blank, Marc & Lebling, Dave (1977). Zork . Cambridge: Infocom.

Bronstring, Marek (2002). "What Are Adventure Games? in

Burden, Donald (1983). Pac-man . Computer game. Atarisoft.

Crowther, William & Woods, Don (1976). Advent(ure) . Computer game. Available for several computer platforms.

Gygax, Gary (1974). Dungeons and Dragons . Board game. Lake Geneva, New York: TSR.

Jones-Steele, John (1987). Tetris . Computer game. Andromeda Software.

Juul, Jesper (2001). "Games Telling stories? A brief note on games and narratives" in Game Studies 1/1, July 2001.

Lewis, David (1973 [revised printing 1986]). Counterfactuals . Oxford: Blackwell.

Manovich, Lev (2001). The Language of New Media . Cambridge, Massachusetts: The MIT Press.

Miller, Rand & Miller, Robyn (1994). Myst . Computer game. Brøderbund.

Newman, James (2002). "The Myth of the Ergodic Videogame: Some thoughts on player-character relationships in videogames" in Game Studies 2/1, July 2002.

Pavel, Thomas G. (1986). Fictional Worlds . Cambridge, Massachusetts and London, England: Harvard University Press.

Romero, John & Carmack, John & Carmack, Adrian (1993). Doom . Computer game. Id Software.

Rosen, Gideon (1990). "Modal Fictionalism" in Mind 99 p. 327-54.

Ryan, Marie-Laure (1991). Possible Worlds, Artificial Intelligence, and Narrative Theory . Bloomington: Indiana UP.

--- (2001). Narrative as Virtual Reality: Immersion and Interactivity in Literature and Electronic Media . Baltimore and London: The John Hopkins University Press.

--- (2001b). "Beyond Myth and Metaphor -The Case of Narrative in Digital Media" in Game Studies 1/1, July 2001.

Schafer, Tim (1998). Grim Fandango . Computer Game. LucasArts.

Searle, John (1975). "The Logical Status of Fictional Discourse" in New Literary History 6; Speech Acts: An Essay in the Philosophy of Language.

Shortt, Alan & Lowe, Jane (1986). Mario Bros . Computer game. Ocean.

Sokal, Benoit (2002). Syberia . Computer game. Microids.

Stern, Gloria. Through the Myst - Another World. A Review of Enter the Myst

Van Looy, Jan (2003). "Uneasy lies the head that wears a crown: interactivity and signification in Head Over Heels" in Game Studies .

Williams, Roberta (1984). King's Quest: Quest for the Crown . Sierra On-Line.








Jan van Looy is Teaching Assistent at the KU Leuven, where he is finishing his PhD on videogames.



Maerlant Center Institute for Cultural Studies

This site is optimized for Netscape 6 and higher

site design: Sara Roegiers @ Maerlantcentrum