The illusion of interactivity
...myth, legend, fable, tale, novella, epic, history, tragedy, drama, comedy, mime, painting, stained glass windows, cinema, comics, news items, conversation. 
The form of the story permeates every aspect of our cultural life. History, politics, memories, even subjectivity, our sense of identity, are all representations in narrative form, signifiers chained together in temporal, spatial, and causal sequence. Narrative is a component of those deep structures with which we construct ourselves and our universe; true stories through which, in the manner of certain Aboriginal legends, the world is dreamed into existence. Narrative appears to be as universal and as old as language itself, and enjoys with language the status of a defining characteristic of humanity and its culture. A people without stories seems as absurd an idea as a people without language, (a people with language but no stories even stranger, for what is language for if not to tell stories?)
Over the past few years there has been a tremendous investment in the idea of digital media, the use of computers as the site of culture rather than just tools for business or science. This is partly due to the drive on the part of manufacturers to create new markets as price/performance ratios in digital technology improve - only recently have cheap computers been capable of simulating analogue sound, stills or moving pictures with sufficient verisimilitude. At the same time, there is a desire at work here, a fantasy which exceeds its technical and economic conditions. Implicit in the notion of digital media is the belief (read desire) that digital computers and digital communications will provide a unified site for 1st world culture in the near future and that this new medium will offer distinct advances over existing media, above all by offering its audience interactivity.
Interactivity refers to the possibility of an audience actively participating in the control of an artwork or representation. Until now, what we call culture has not allowed for a great deal of interaction from the audience. The audience is given a space for interpretation and a space for reaction, but not for interaction. There are those who argue that interpretation is interaction, and so of course it is, but not in the sense intended here. For the purposes of this discussion, interactivity means the ability to intervene in a meaningful way within the representation itself, not to read it differently. Thus interactivity in music would mean the ability to change the sound, interactivity in painting to change colours, or make marks, interactivity in film the the ability to change the way the movie comes out and so on. In its most fully realised form, that of the simulation, interactivity allows narrative situations to be described in potentia and then set into motion - a process whereby model building supercedes storytelling, and the what-if engine replaces narrative sequence.
There are those who see the replacement of narrative form by interactive simulation as political progress. Many who in the 60s and 70s rejected the blandishments of mainstream narrative, the elision of its own means of production and the naturalisation of passive spectatorship, discern in interactive media an opportunity to go beyond the impasse of avant garde structural materialist film practice (see Zap Splat by Malcolm leGrice) Similarly, in the rhetoric of neo-liberal political thought interactivity can be figured as a form of freedom, a liberation from the tyranny of authorship and the servile passivity of reading. Interactivity in this context is something that gives the individual more choice, takes the mediation out the media and empowers us all as free consumers in the marketplace of signs. Interactivity here is something to believe in, a democratic revolution in representation...
This discussion is an attempt to speculate on the collision between a dominant cultural form - narrative, and the technology of interactivity. I will argue that there is a central contradiction within the idea of interactive narrative - that narrative form is fundamentally linear and non-interactive. The interactive story implies a form which is not that of narrative, within which the position and authority of the narrator is dispersed among the readers, in which spectatorship and temporality are displaced, and in which the idea of cinema, or of literature, merges with that of the game, or of sport. The consequences may be far-reaching and profound. Can an interactive construct, or a simulation, successfully adopt a narrative form? Will there be a general transformation from a culture of stories to a culture which expresses its truths through an immersive, interactive medium, the shared experience of the simulator?
Forking Paths and Synthetic Spaces
In his short story Garden of Forking Paths , Borges imagines a novel in which the path of the story splits, where all things are conceivable, and all things take place. The author of this story within a story is judged insane and commits suicide, and Borges’ narrator is arrested and condemned to death - thus the fate of the narrator and of the author in the interactive era is prefigured. It is not hard to see how the task of writing interactively might drive an author to insanity and suicide. To write not simply an account of what happened but a whole series of ‘what-ifs’ increases both the volume and complexity of an author's task exponentially. And if the reader chooses his or her own pathway through the story then the narrator - or director - can be done away with; in effect the function and authority of authorship is usurped by the reader. The ability to shout ‘cut’ or to make a gun go off, or to develop the action in a particular direction is no longer the unique perrogative of an omnipotent author but is partially devolved to the audience.
Interactivity implies forking paths and each pathway must be written and fitted together. The greater the number of pathways, the greater the sense of textual play for the reader, and the greater the amount of work for the writer. The volume of story web increases exponentially with additional points of interaction. An author is faced with an inevitable and depressing tradeoff - sacrificing time spent on the texture of the narrative, its literary or cinematic qualities, for an enhanced interactive complexity. The result can be interactive but schematic, resembling the outline of a story rather than the story itself.
How much interactivity does it take to make an interactive story? We don’t know because we don’t know what an interactive story is like, nor what it is for (more on this in a moment). It is true that the number and complexity of forking paths could be increased until the reader experiences a large degree of freedom and control within the text. The limits of this freedom are achieved within a model that dispenses with the network of lines altogether, replacing it with a fictional space within which readers can turn left or right, look up or look down, open a door, enter a room, at any time they choose - a spatio-temporal simulation which can generate a travelling point of view in real time, more commonly known as virtual reality or VR. In the VR model, although the reader/spectator enjoys seamless temporal and spatial liberty, the tradeoff between interactivity and richness of content holds true. VR to date has barely been able to dress the set, let alone cry ‘action’, or murmur ‘once upon a time’.
If the sheer complexity of building an interactive narrative is problematic at the technical level, there is another simpler and deeper problem. This is the question of what kind of representation an interactive representation is, if you like, the question of ontology. The change from a linear model to a multi-linear or spatio-temporal(VR) model is more than just the change from a simple line to a more complex diagram or model; it involves moving from one kind of representation - and one form of spectatorship - to another.
A Lonely Impulse of Delight
As he settled into the snug cockpit he tried not to think about the obvious thing. Ahead of him, through the windscreen, he could see a long low hill. It was further away than it appeared to be, and much bigger. Yellow through the blue haze, the hill squatted on the plain, low and indolent and massive. He wanted to be over that hill and look beyond.
Before him stretched the grey runway, on the left a yellow haystack, on the right a white airfield building. All around him was the blue airplane.
He opened the throttle and the plane began to inch forwards. The nose veered to the right, towards the white building, and he rapidly trimmed the plane to the left.
By now the ground was rushing past and the tail starting to lift. The nose came down and he could see the ground immediately in front for the first time, a streaming grey blur, and the end of the runway rushing up to meet him. At the last possible moment he pulled the stick back into his stomach and the plane lurched into the air. Vertigo.
Afficionados of the Hellcats flight simulator will recognize the landscape - an American airstrip on one of the Solomon Islands in the Pacific Ocean. The time is WW2. This is the prologue to an account of an experience of my own, flying a Hellcat on a mission against the Japanese Navy.
Hellcats is effectively a screen and mouse based virtual reality system - 2nd person VR - offering non-linear adventure stories. The reader - or should it be participant, or player - is free to move in any direction, at all times, as long as he or she never gets out of the plane. This cuts down the scope of the story significantly - it's like Top Gun with everything but the flight scenes cut out. Hellcats is a simulator which models a space and a set of rules - the aerodynamics of a propeller plane - for moving through that space. It provides a simple narrative framework within which to act - the struggle against the enemy - and it provides characters to interact with and who appear to be independent narrative agents with their own characteristics and motivation - Japanese pilots, gunners and sailors.
As a representation of the experience of Americans during WW2 in the Pacific, Hellcats can be compared to South Pacific or From Here to Eternity. Yet despite the similarities of place and time, Hellcats is a very different kind of representation. Hellcats represents one specific aspect of the experience of the war in the Pacific, but it is the experience of the machine, to misquote Stephen Heath, rather than the experience of the pilot. More precisely, it is the experience of the pilot insofar as he or she is an extension of the machine, that part which keeps the plane in the air and flies this way or that way, presses the trigger and drops bombs, but never that part with a history, a family, skin colour, memories, desire, plans for this evening...
Certain key attributes of narrative form are missing.  Narrative closure has to be fought for - if you crash your plane while taking off the ‘story’ is short, insignificant and unsatisfying. It is up to the spectator to ensure that the action comes to a satisfying and meaningful end - closure is not part of the structure of the representation but is contingent on the moment of ‘reading’. Temporal and spatial coherence are more or less complete, but strictly limited to the skies above the Solomon Islands. There is no specific enigma to be resolved but a different kind of teleological imperative, that of a participant in a violent struggle. If we consider what Barthes has called the symbolic code, that code which accounts for the formal relationships created between terms within a text - the figurative patterning of antithesis, graduation, repetition etc, we find it absent in Hellcats. The simulator does not signify in this way. Neither do we find much in the way of a referential or gnomic code, the code of shared cultural knowledge about the world, nor the rich and diffuse code of connotations designated by Barthes as the code of semes. What is lost is the complex interplay of signs, Barthes’ ‘weaving of the voices’ across different registers, the ‘perspective of quotations’, the ‘mirage of structures’, the ‘multivalence of the text’. These are replaced with a wide band of sensory information refering to specific and schematic aspects of a situation - the proairetics of flight, the hermeneutics of battle. 
However, although complex narrative codes are not hard-wired into the simulation, they are not therefore altogether absent from it. The simulation is re-invested with narrative sense via the subjectivity of the participant - as if subjects have a will-to-narrative which imposes its order even in the sparsest of contexts. This is a narrative which issues from the identifications and interpretations the participant makes during the interactive experience - a personal, transient and contingent narrative unlegitimated by the external figure of the author.
I saw the movie last week. I want what happened in the movie last week to happen in the movie this week too, otherwise what is life all about? 
A key distinction to be made between an interactive representation, like Hellcats, and narrative representations like those of the cinema and literature, lies in the way time is represented. Narrative refers to the past. It is an account of events which have already taken place. Its temporal referent is once upon a time. This relationship to time is not affected by the verb tense - the present tense is often used to bring immediacy and drama to an account - nor does it depend on the reality of the events being described - fiction gives an account of things which happened, which is nonetheless untrue. The historical quality of narrative appears to be part of its very nature as representation, its ontology. The simulator on the other hand operates in the present. If in a narrative an event happened, in an interactive narrative, whether multi-linear or spatio-temporal, an event is happening, its time is now. This temporal shift has important consequences.
A linear narrative exercises a textual authority which is dispersed by interactivity. In a linear narrative, the reader submits to the prior authority of the text. Only the author has the power to make decisions about the story-line or point of view, and the invention of narrative sequence is his or her sole perrogative. The text is certain of itself. Moreover this certainty has a legitimising function. Hayden White writes:
‘We cannot but be struck by the frequency with which narrativity, whether of the fictional or the factual sort, presuposes the existence of a legal system against or on behalf of which the typical agents of a narrative account militate. And this raises the suspicion that narrative in general, from the folktale to the novel, from the “annals” to the fully realised history, has to do with the topics of law, legality, legitimacy or more generally “authority”. 
Now this authority is expressed, and legitimacy conferred, at the moment of closure. By recounting what happened an author is also closing of those things which didn’t happen. A character picks up the phone rather than letting it ring, someone walks down the street and turns left instead of right. Closure in this sense is dispersed throughout the narrative. The events unfold as a pattern which progressively resolves itself into an image, each event integrating those which precede it into progressively higher levels of narrative sense. This process continues until the final closure at the end of the narrative, at which point the meaning of the story is revealed at last, and is revealed to have been immanent in all the events all along. Closure can be considered as a function of time, or more precisely of the way in which time is represented, whether as past and complete or present and ongoing. 
In his standard work on aspect , the linguist Bernard Comrie distinguishes two forms of time reference in language - aspect and tense. Tense ‘relates the time of the situation ... to some other time, usually to the moment of speaking’, whereas aspects are ‘different ways of viewing the internal temporal constituency of a situation’. Where tense distinguishes between situations taking place in the past, present or future, aspect draws a distinction between the perfective ; a situation viewed from the ‘outside’ as completed, and the imperfective; a situation viewed from the ‘inside’ as ongoing. The shift from narrative representation to interactive representation entails an aspectual shift like that from perfective to imperfective, from outside to inside the time of the situation being described. Narrative representation can thus be considered as analogous to the perfective aspect - always already completed by closure, apprehended as a totality as if seen from a distance and without the possibility of change, whereas interactive representation is analogous to the imperfective aspect - incomplete, seen from the inside with all the messiness and incoherence that this implies, ongoing.
It is important to recognize that aspect does not distinguish between different types of situation but between different ways of representing a situation, or more precisely between different ways of positioning the audience with respect to a situation. This suggests that perfective and imperfective aspect, and by analogy linear narrative and interactive simulations, correspond to two fundementally different modes of spectatorship.
If a story refers to a chain of events that have already taken place, that have been completed in some sense before the story begins (otherwise how could one tell a story about them?) what might an interactive story or simulation refer to? An interactive simulation appears to designate the conditions for events rather than the events themselves. The interactive simulation sketches a web of possibilities and constitutes a system for producing story events in time - a story engine. Closure - the cutting out and sequencing of events from the mass of possibilities - is effected by the spectator, albeit within a framework of conditions and possibilities designed by the author.
It is in the differing aspect of their respective modes of closure that we can locate the apparent disjuncture between the nature of interactivity and that of narrative. The moment the reader intervenes to change the story (at the nodes of multi-linear narrative or at every moment in a spatio-temporal simulator) is the moment when the story changes from being an account of events which have already taken place to the experience of events which are taking place in the present. Perfective becomes imperfective, story time becomes real time, an account becomes an experience, the spectator or reader becomes a participant or player, and the narrative begins to ressemble a game.
So far I have argued for a distinction between narrative and interactivity, or stories and games, which is based on the different way each represents time, leading on to differing modes of spectatorship. However, as Max Whitby points out, games and stories also have very different cultural values attached to them. The game is frivolous whereas narrative is serious - the form of the game is agonistic and ephemeral, dealing in a form of transient athletic display. The game is an exercise which can exist only in the present (if it persists in memory then it does so as an account).
There is a general assumption here that narrative representation - literature, history, cinema and so on, has a deep and lasting significance which the game lacks. In the end Shakespeare or Proust or Pasolini seem to have more to offer than a game of football or Sonic. The game is outside of history, unworthy of serious remembrance. At the M.I.T. multimedia conference in Dublin in 1993 a speaker bemoaned the fact that his son spent too much time playing computer games and not enough time reading books. Thinking of my own child, I found myself nodding in agreement. Yet when a woman asked from the floor why reading a book was better than playing a computer game, he couldn’t explain his assumption and neither could I. Two other speakers gave a fascinating account of an elastic movie. This was a multi-screen installation constructed as part of a student workshop at M.I.T. which the spectator moved through and interacted with. The speakers called it an interactive media environment, an installation, a transformational space, fine art circumlocutions for the obvious term game which they managed to avoid entirely throughout their paper. Then they showed a video of their undergraduate students discussing the design of the project and the word game cropped up over and over again. Finally, throughout the whole 2 day conference on interactivity, discussion of console and TV computer games was almost entirely absent, in spite of the astounding commercial success of Nintendo and Sega in the youth market, in spite of CD-i, in spite of 3DO...
In my class in interactive media at the University of Westminster I encourage the students to play computer games in order to give them a sense of the possiblities - and limitations - of the crossover between interactive representations and the story. This did not initially meet with the approval of the department and there is still a lingering suspicion that those students who take the module in interactivity just want to play computer games. Yet nobody accuses the film students of just wanting to watch films. Many college computer rooms have a notice on the wall warning that the playing of games is banned. The game is not work but a diversion from work, nor is it a proper object of serious study. The game is something which, although tolerated, the law must seek to repress, to keep to its proper place.
A Literary You-topia
If the repressed reading of interactivity is that of the game, the preferred readings are interactivity as liberation, and interactivity as Post-Modernism come true.
In S/Z, Barthes describes two types of writing, readerly writing and writerly writing. What happens if we take the notion of the writerly at face value, innocently? Let us reproduce the notion of the writerly - or rather, let us post-produce it. Let us abolish the distinction between the producer (Barthes) and the reader (me, you) and rewrite the writerly. Let us read excessively, irresponsibly, futuristically.
‘The goal of literary work (of literature as work) is to make the reader no longer a consumer, but a producer of the text...
The writerly text is a perpetual present, upon which no consequent language (which would inevitably make it past) can be superimposed; the writerly text is ourselves writing...
In this ideal text, the networks are many and interact, without any of them being able to surpass the rest; this text is a galaxy of signifiers, not a structure of signifieds; it has no beginning; it is reversible; we gain access to it by several entrances, none of which can be authoritatively declared to be the main one; the codes it mobilizes extend as far as the eye can see ...’ 
In this excessive reading the writerly becomes a fantasy of the multi-linear text, Barthes a kind of Nostradumus of literary theory, writerly writing the uncanny prophecy of an interactive literature come to pass. Indeed, a number of commentators have noted the way in which poststructuralist writing seems to anticipate the non-linearity of new technology. In Hypertext - the convergence of contemporary critical theory and technology, George P Landow suggests that the literary theories of structuralist and post-structuralist thinkers (especially Barthes and Derrida) find their embodiment in interactive hypertextual forms made possible by new technology. Hypertextual and non-linear structures promise Barthes’ writerly text, never far from the possibility of rewriting, multivocal, decentred, without boundaries, a text which can break free from the chains of closure, a text whose instability lies not in our postmodern apprehension of it but in its very condition of being. Hypertext for Landow is post-structuralism made flesh, transubstantiated - Foucault’s death of the Author a corpse and a smoking gun, Derridean débordement actualized as hypertextual annotation... 
The problem with this kind of literal and utopian mapping of post-structuralist theory onto new technology is that it fails to acknowledge its own excessiveness. To literally and deterministically locate a set of complex, heterogenous and ambiguous ideas about the social processes of reading within a specific technology seems to be missing the point. One might as well argue that the telephone system is post structuralist. It is ironic that a set of theories which stress plurality and indeterminacy should be employed in the service of a reductive equivalence between very different types of object, a critical discourse of interpretation on the one hand and a machinery of interaction on the other. Just as theory is not praxis, interpretation is not interaction.
‘Science has always been in conflict with narratives’ 
We have seen how a putative theory of interactivity might oscillate between the preferred register of the post-modern (serious, plural, decentred and legitimated by the academy) and the frivolous register of the game (playful, ephemeral, banal and without value). A further approach is suggested in The Postmodern Condition in which Lyotard outlines an opposition between narrative knowledge (convivial, traditional) and instrumental knowledge (cybernetic, scientific). The game can be considered as a cybernetic construct (a goal directed system of control and feedback) and as such, placed on the side of the instrumental, whereas narrative knowledge, argues Lyotard, is an older form - ‘narration is the quintessential form of customary knowledge...’ and ‘what is transmitted through narrative is the pragmatic which establishes the social bond’. Legitimation and authority are immanent to narrative form and are established within and through the act of narration itself (see Hayden White quote above). By contrast authority and legitimation are extrinsic to the form of instrumental knowledge. In scientific discourse legitimation must be fought for. Moreover, instrumental knowledge according to Lyotard is set apart from the language games that constitute the social bond. The analagous oppositions may be summed up thus:
These oppositions sketch out the structural differences between two different kinds of representation, and two modes of spectatorship. It seems that the truth-effects of stories and games are very different.The question of legitimacy and certainty is central - the simulation remains a model which does not have the ability to auto-legitimate itself in the way an account does. Structured as it is around a core of what-if statements, the truth of a simulation or game can never be more than hypothetical. 
Interactive Critical Theory
How then to approach the question of a critical theory of interactive representation? We might start by looking at early prototypes of interactivity on CD-Rom and laserdisk. It is vital here to insist upon the distinction between the multi-linear (Forking Paths) and spatio-temporal (VR) models. In a multi-linear construct the author can play with the space between linear sections - versions of what happened, or different points of view, connected within an authored network of simultaneity and sequence. What is explored here is the space between alternative sections of writing or video - the space is a properly literary or cinematic space rather than the cybernetically governed mechanical space of VR, and one can imagine this hyperliterature being opened up to an expanded literary criticism.
The author of a simulation or VR representation works with a different set of opportunities. Here it is not so much a question of writing in space but of designing a model. Although the author describes the characteristics of the model, he or she is not the author of the events that happen within the model once set in motion. Here, as I have already discussed, it is more difficult to talk of authorship at all. An as yet unformulated critical approach to the simulation will probably be informed by cybernetics, architecture and the theatre. 
A number of early forays into interactive media can be considered as prototypes for multi-linear writing and as pointers to a criticism of new cultural and literary/cinematic forms. These demonstrate a tension between repression and freedom, offering the reader the illusion of control within a tightly authored set of possibilities. The multi-linear model has the advantage of being based upon and incorporating an older model, that of linear writing - a model grounded at least in part within the narrative tradition, although exceeding and threatening that tradition at the same time.
In Graham Weinbrene’s interactive cinema piece Sonata, the viewer is offered control over the aspect16 of the narration - the screen is divided into four temporal regions, left for flashback, right for the present, up for an expanded present and down for filmic elements which are outside of the time of the story altogether. In Tolstoy’s original short story the narrator unburdens himself to a stranger on a train, telling how, consumed with jealousy over an imagined affair between his wife and her music teacher, he knifed his wife to death. In Weinbren’s version the viewer is able to control the flow of narration and view the events either as perfective - seen from within the time frame of the events, or imperfective - from the external vantage point of the future. 
Thus if you point at the right of the screen you get the murderer recounting his story in the railway carriage, and if you point at the left you get the dramatic events played out in flashback. The sequence of events represented by Weinbren stays the same, however the mode of telling can throw the spectator inside or outside of those events. By pointing up or down you can overlay the fevered imaginings of the jealous husband (a sex scene between wife and music teacher), the mouth of the wife of Tolstoy cursing her husband's misogyny, references to Freud’s The Wolf Man and the classical image of Judith with the severed head of Holofernes. The climax of the piece is an interactive wipe which the spectator controls by waving a finger at the screen - outside the music room the agitated husband paces up and down while inside the wife and teacher practise the Kreutzer Sonata unaware of the tragedy about to befall them.
The experience of viewing Sonata is both exhilarating and dislocating. Unlike a fully interactive fiction in which story events themselves are switchable, Sonata progresses inexorably from beginning to bloody end, but the route taken is profoundly different with each viewing. One showing might be as grammatically correct as a costume drama on the BBC, another as obliquely avant garde as a French art movie. The interactivity here doesn’t ‘get in the way’ as Max Whitby suggests, but provides an extra dimension within which to write and read the movie.
Claudia Frutiger, Alejandra Jiminez and Kate Reddit have recently authored in London an interactive eternal triangle in which 3 strangers, thrown together for the night in an isolated hotel, ponder which of the other two they can bear to share the only room with. The story offers the viewer the chance to control their identifications with the characters - by choosing a character’s point of view, that character becomes the protagonist around which the story organises itself. Each point of view is partial - what is concealed from each character is more important than what is revealed. The story is cyclical, complex, enigmatic and without resolution.
What these experiments reveal is a tension between gameplay and the story, between the instrumental and the narrative function. To put it simply, the more of a story it is, the less of a game, and vice versa. A reconciliation of this impasse suggests itself from a surprising quarter. The form of pornography is both narrative and game-like, referring to a sequence of fictional events and a kind of arousal game with a clearly defined outcome - sexual pleasure. The pornographic story joins the reader in a cybernetic construct - within this cyborg-text the body of the reader and the body of the text respond to each other. Pornography has been well represented in early interactive commercial products and the notion of virtual sex is the dominant popular fantasy about VR (at least among journalists).
There are two potential endings for a discussion like this, either optimistic or pessimistic. Neither is sustainable. The ‘interactivity is post modern’ school of thought sees interactive representation as a liberation from the repressive authority of traditional narrative form. There are echoes here from the avant-garde and anti-narrative movements in cinema and writing which have their source in the utopian ferment of the 60s. (See Zap Splat by Malcolm le Grice) Yet the consequences of the opening up of closure - that interactivity will be ‘commonplace, unlaborious, shallow, un-literary, heterodox’  are more difficult to accept.
Others see the simulation as promising post-symbolic representation, bypassing the patriarchal distortions of perspective and the controlling point of view. An interactive simulation, according to this argument, offers not the representation of objects but the representation of relations between objects within which the participant can select their own point of view. By using immersion interfaces the participant can gain, so the argument goes, direct (ie unmediated, objective) access to pure data, (a realm both digital and noumenal). However, in characterising this as a shift from coded representation to experiential post-representation what is glossed over is the coding and mediation involved in constructing the simulation in the first place. Sim City, the town planning simulation game, is just as much a cocktail of opinion, received wisdom and poliitcal ideology as any other doctrine of urban decay and renewal - it simply hides its politics more effectively. These are old arguments against the naturalisation of representation, but it seems that they will need to be rehearsed once again.
Is this the end of the road for narrative, grand or otherwise? Are we to become a people without stories? Once again the linguistic category of aspect provides a useful analogy here. We have seen how the shift from narrative to the interactivity involves a shift from perfective to imperfective, from outside to inside the time of the events being described. Thus narrative representation and interactive representation might be ‘different ways of viewing the internal temporal constituency of a situation’ as well as different forms of spectatorship.  As interactivity increases, so the spectator is thrown inside the representation to become a player. This increasing involvement of the spectator within the frame needs to be situated within the broader context of a long historical trend within the visual arts, and most recently within the development of film language, towards techniques of encouraging identification between viewer and artwork. But if an interactive simulation can almost literally put you in the frame, it does so at the expense of narrative coherency - unlike film, interactive media has not so far developed a language which allows identification within a narrative framework.
At the heart of the interactive representation narrative reinstates itself through the subject narrativising the experience, making sense from (simulated) events. If narrative is a technique for producing significance out of being, order out of contingency, then simulation can be seen as its inversion, a technique for producing being out of significance, of generating a simulation of contingency from first principles. Rather than a people without stories, interactivity offers the promise of a people within stories, and rather than the end of narrative, an explosion of narrative within the simulator.
Like any other form of representation, interactivity is an illusion. It puts itself in the place of something that isn’t there. What then might be the absent referent of interactivity? According to both neo-liberals and techno-utopians interactivity promises the spectator freedom and choice. It is precisely the absence of such freedom and choice that interactivity would appear to conceal.
 Roland Barthes, 'The Structural Analysis of Narrative' in Image, Music, Text
 Jorge Luis Borges, Ficciones
 Annette Kuhn in The Cinema Book, British Film Institute 1987, gives this account of the formal attributes of classic cinematic narrative:
* Linearity of cause and effect within an overall trajectory of enigma-resolution;
* A high degree of narrative closure;
* A fictional world governed by spatial and temporal verisimilitude;
* Centrality of the narrative agency of psychologically-rounded characters.
 In S/Z, Blackwell 1990, Barthes outlines 5 codes of narrative. These are used to submit a short story by Balzac - Sarrasine - to an extremely close textual analysis. Briefly the 5 codes are:
* The code of Semes - broad connotations within the text - femininity, age, etc...;
* The Symbolic code - the code which structures the text in figurative patterns - antithesis, graduation, repetition etc. It is difficult to imagine this code within a non-linear, interactive structure - the pattern imposed by the author would be lost in the meanderings of the reader;
* The Cultural code - shared knowledge, common sense. See note on Hellcats above;
* Hermeneutic code - ‘the various (formal) terms by which an enigma can be distinguished, suggested, formulated, held in suspense and finally disclosed.’ An interactive story might be organised principally in terms of the hermeneutic code, a cluster of clues surrounding a mystery could be organised logically yet non-sequentially. The hermeneutic code is goal-oriented, as are most games;
* Proairetic code - the code of the actions, the code of the sequence. This code presents particular problems for non-linear interactive structures. A change in one part of the sequence will have the potential to change every subsequent action. The proairetic code embodies a relentless logic - if X is killed in scene 4 then X cannot be alive in scene 5. To an extent then, the proairetic code embodies something of the Cultural codes, the code of knowledge. The proairetic code is the code of knowledge about time, and it is the certainties of this knowledge which interactivity appears to throw into question. There is a parallel between the interactive narrative and the electronic spreadsheet. The linear narrative is to the interactive narrative what the ledger is to the spreadsheet. Both interactive narrative and spreadsheet are ‘what if?’ engines. Both create the space for multiple parallel time. The best illustration of the problem of the proairetic code in interactive narrative is given by changing one of the numbers in a spreadsheet, doing a recalculate, and watching the changes multiply and ripple across the whole sheet.
 Woman outside the cinema in The Purple Rose of Cairo directed by Woody Allen.
 'The Value of Narrativity' in The Representation of Reality: On Narrative, Chicago, 1981.
 See Peter Wollen’s discussion of the linguistic category of aspect and its effect on spectatorship in ‘Fire and Ice’ in Richon & Berger (eds.) Other Than Itself, Cornerhouse Publications 1990 and the categories of perfective and imperfective in Bernard Comrie’s standard work Aspect, CUP 1976.
 Bernard Comrie,Aspect, CUP 1976
 'Herbert Quain' in Jorge Luis Borges, Ficciones.
 Max Whitby heads the Multimedia Corporation, an offshoot of the BBC which produces interactive titles in London.
 Roland Barthes, S/Z, Blackwell, 1990.
 ‘contemporary theory proposes and hypertext disposes; or, to be less theologically aphoristic, hypertext embodies many of the ideas and attitudes proposed by Barthes, Derrida and Foucault.’ Landow, Hypertext, Johns Hopkins University Press 1992, page 73.
 Lyotard, The Postmodern Condition, Manchester University Press 1986.
 For example, my 5 year old child enjoys crashing the aeroplane when he flies the simulator - it doesn’t hurt him to crash the plane. However when watching a television documentary about early USAF jet planes which showed a plane cartwheeling and exploding in a fireball he was upset because he felt he had seen someone die. The simulated crash and the account of a crash had for him a very different status.
 See Brenda Laurel, Computers as Theatre.
 Linguistic category distinct from tense. See footnote referring to Comrie, Aspect, CUP 1976.
 ‘Another way of explaining the difference between perfective and imperfective meaning is to say that the perfective looks at the situation from the outside, without necesarily distinguishing any of the internal structure of the situation, whereas the imperfective looks at the situation from the inside, and as such is crucially concerned with the internal structure of the situation, since it can both look backwards towards the start of the situation, and look forwards to the end of the situation, and indeed is equally appropriate if the situation is one that lasts through all time, without any beginning and without any end.’ Bernard Comrie, Aspect, CUP 1976.
 'Herbert Quain' in Jorge Luis Borges, Ficciones.
 Holt's defintition of aspect, quoted in Comrie, Aspect, CUP 1976./note