I recently retrieved my old dissertations from MA Design for Interactive Media course back in 1995… This one discusses the concept of interactive narrative fiction and the problems faced in attempting successfully to create such works. Re-reading it now, seventeen years later, I am particularly struck by the relevance of this excerpt to so-called ‘Interactive’ (i.e. ‘red-button’) TV:
“It seems to me that creating interactive narrative is not just a matter of taking conventional prose and putting it into an interactive medium. Indeed, it might be argued that there is more to be lost from doing this than there is to be gained.”
Probably best if I don’t get started on what’s wrong with “interactive TV”… ;-) But I do have another old article on that lying about somewhere, which I may dig out and post up here at some point. Anyway, hope you enjoy this one.
Interactive Narrative Fiction
Submitted in partial fulfilment of module ADC4210 of the MA Design for Interactive Media, Middlesex University
3rd February 1995
Abstract
This essay is a discussion of approaches to interactive narrative fiction, and the problems inherent in creating such works. It focusses in particular on tree fiction and the open work, the merging roles of writer and reader, the notion of interactive narrative as adventure game, and some commercial considerations. It is suggested that, at present, most examples of interactive narrative are not particularly successful – in particular they are not satisfying to read. It is further suggested that an important obstacle to their success is that they fall somewhere between the novel and the adventure game, but fail to fulfil the requirements of either readers or players. However, the most significant problem may be that we, as readers, are not yet familiar with the modes of thought necessary in order fully to appreciate this new medium.
Contents
Introduction
Approaches to interactive narrative
Hypertext and conventional narrative
Hypertext as narrative structure
Tree fiction
The open work
The writer and the reader
Games and narrative
Commercial considerations
The project
Conclusion
Bibliography
Introduction
“Now a whole is that which has beginning, middle, and end. A beginning is that which is not itself necessarily after anything else, and which has naturally something else after it; an end is that which is naturally after something itself either as its necessity or usual consequent, and with nothing else after it; and a middle, that which is by nature after one thing and also has another after it.” (Aristotle, on narrative, Poetics)
A narrative is a story, an account of something either real or fictional, which has `happened’ in the past. It has a number of important defining features, in particular, closure. By Aristotle’s definition, a narrative must have a beginning, a middle and an end; the end occurs at the moment of closure, when the sequence of events have led to their conclusion, when the meaning of the story is revealed. There is also closure throughout the narrative as, by recounting what did happen, the author closes off that which did not take place, and it is this closure which lends the narrative its authority. Only the author can make decisions about the plot, events, point of view and so on; the reader can only submit to the authority of the text or stop reading (Cameron 1993).
Interactivity, arguably, changes all that. It allows the reader to be actively involved in the creation of the narrative, to choose where to start and finish, what to see, what not to see, which events happen, which don’t. As such, it questions Aristotle’s ideas of plot and story. As Landow (1992) points out, this either means that we cannot write interactive fiction, or Aristotle’s ideas don’t apply to such works. So, the question is can successful interactive narrative fiction be written?
Approaches to interactive narrative
Most approaches to interactive narrative are based around two concepts in particular. Both have been greatly facilitated by the computer which seems to have become an important platform for the development and use of such works – although it is possible to create interactive narrative in other media, it would seem that most are now computer-based. The two concepts are the hierarchical, multi-linear tree structure, and the open, fluid, anti-linear web structure. Indeed, they are brought together to a significant extent in the adventure game, in which the player typically moves through a virtual environment (the web structure), interacting with objects and characters and attempting to fulfil some objective. These interactions may lead the player down particular paths of the adventure’s plot, determining which events happen and which don’t (hence the tree structure also). In the process, it might be said, the player is taking part in an interactive narrative. However, while the adventure game is perhaps the most popular form of computer-based interactive narrative (albeit largely due to a dearth of anything else – and this may be for good reason as will be seen later) it is not the only one.
Before proceeding further, an important point must be discussed. Although interactive narrative seems to be based predominantly around non-linear or multi-linear structures, this does not mean that conventional narrative is totally linear. While any one reading of any narrative, be it conventional or interactive, will be perceived by the reader as being linear, there are many ways in which authors of conventional narrative can alter the linearity of the plot or story. I take `plot’ to mean the actual sequence of events in the narrative, and `story’ to mean the order in which they are told to the reader.
A linear plot may be presented in the same linear sequence, so that the events in the story are experienced by the reader in the order in which they occur. However, even here, the author can play with the reader’s perception of the time involved – a very brief event (for example, a car crash) may be stretched out over several pages; a very long series of events, perhaps over several years leading up to the events of the story, may be summarised in a brief paragraph. Alternatively, a linear plot may be presented in a non-linear sequence, through the use of flashbacks, flash forwards and so on. An example of this might be Quentin Tarantino’s film Pulp Fiction in which a temporally linear series of events is shown in a non-temporal sequence.
In either of the above cases, there may also be one or more additional plots interwoven into the story; these may be occurring at the same time, in the same sequence, or in different times and different sequences. Further, it is conceivable that even the plot may be made non-linear, although if we accept that a plot consists of a “chain of events in cause-effect relationship occurring in time and space” (Bordwell and Thompson, 1990), then we may run into problems of backward causation and such like.
The point which I hope is clear from this is that there is plenty of scope for non- or multi-linearity in conventional narrative. All this also applies to interactive narrative, but what distinguishes it from the conventional variety is that its presentation may also be non- or multi-linear. In the conventional narrative, the pages are numbered in linear sequence; one normally starts from the beginning and works through to the end. Although the plot and story may not be linear, all are defined by the author and are unalterable by the reader. In the interactive narrative, the sequence of pages, or other units, is determined by the reader; the consequence of this is that, in contrast to conventional narrative, each event is not necessarily in any particular context – its context may be different on successive readings.
Hypertext and conventional narrative
On the computer, most interactive narratives are realised using hypertext. There are a number of ways in which hypertext can be added to conventional narrative. It may be used, for example, simply to add further information – a hot spot may lead to more information about a character or place, or to a footnote or illustration. Or it might be used to access the different chapters or sections, as in the hypertext adaptations of Peter James’ Host and Martin Gardner’s The Annotated Alice. The problem with this, however, is that it begs the question – why is it really necessary? While this use of hypertext undoubtedly has many academic uses – for example finding out if a particular character appears in any other works by the same author – it does not seem to offer much benefit to anyone reading purely for pleasure, especially since the computer screen is not a particularly comfortable reading medium. Granted a hypertext contents page saves time by removing the need to flick through pages in search of a particular section, but anyone who tries to read more than a few hundred words on screen may soon come to accept this as a mixed blessing.
In addition, reading such material is not particularly fulfilling. The interactive story The Doomsday Brunette for example contains a number of pieces of hot text which lead to further details about their subject. I personally found the temptation to click on all the hot spots distracted me from the story and diminished my sense of involvement. On the other hand, not clicking on the hot spots left me with a feeling that I might have missed something. If the text arrived at from a hotspot is not relevant to the narrative, then why include it; if it is relevant then why not include it in the main body of text?
It seems to me that creating interactive narrative is not just a matter of taking conventional prose and putting it into an interactive medium. Indeed, it might be argued that there is more to be lost from doing this than there is to be gained. I feel there are few conventional narratives which can successfully be translated to this medium, although Landow (1992) suggests Finegan’s Wake, Tristram Shandy, In Memoriam and Ulysses as possibilities. It would seem, though, that for the most part successful interactive narrative is more likely to be written specifically for this purpose.
Hypertext as narrative structure
In this section I shall discuss some of the approaches which may be taken to designing interactive narrative from scratch.
Perhaps the most common form of interactive narrative at present (with the possible exception of the adventure game) is that based around a hierarchical, multi-linear tree structure. Such works comprise short sections of narrative each of which is followed by a node, or decision point where the reader chooses what happens next from two or more options. Printed examples of such narrative can be found in the “Choose Your Own Adventure” series, or for electronic examples see Stories from Downtown Anywhere, Addventure, Accounting for the Cards or The Doomsday Brunette on the World Wide Web .
Such structures allow the reader to explore different possible events within the narrative, to reach alternative outcomes, to answer all the “what-ifs” inherent in any story. While this may seem a very natural idea (Rees, 1994), it does lead to a number of problems.
In particular writing a narrative which not only tells one story but also all (or even just some) of the alternative possibilities leads to an exponential increase in the volume and complexity of the work (Cameron, 1993). If a narrative, starting from a single scene, were to have just four binary decision points (i.e where there are only two alternatives to choose between), then the whole work would contain a total of thirty-one scenes. Rees (1994) points to Alan Ayckbourn’s play Intimate Exchanges (1985) as an example of such a narrative; he also notes that it fills “two weighty volumes”. With ten binary decision points, such a narrative would have over two thousand scenes; with twenty, more than a million. If any of the decision points had more than two options (as many electronic hypertext narratives do – eg The IMRF Choose Your Own Adventure Story ) , then clearly the number of scenes would increase still further.
So, if the narrative is to be substantial enough to give the reader time to become involved in it, how is the author to overcome the sheer amount of time it would take to create such a piece? Rees (1994) suggests two possibilities.
One solution is to reduce the number of scenes by merging narratives – in other words by making several paths lead to the same place. This gives the illusion of greater complexity than really exists. But surely this rather defeats the point of writing a tree-structured narrative? If two (or more) paths lead to the same outcome, then the reader, in choosing between them, has not really made a choice at all. What if something significant happens on a path which then leads to a point which could also have been reached via a different path on which that event did not happen? Any consequences of that event have been lost – it might as well not have happened, the first path need not have been there, there was no need for the reader to make a choice. This assumes, however, that the outcome is all that matters, which may not be the case. Hence this situation may not be as bad as it sounds. It might be overcome by careful consideration on the part of the author as regards which paths can be successfully merged. The story need be no less entertaining or engaging, and there is not (of necessity) any loss of consequence, if the merged paths are only those which contain relatively trivial or insignificant information (for example, the protagonist goes from one place to another; in one path he travels by train, in another, by car). The important thing is for the consequences of all significant events to be fully followed through. Alternatively, the use of merging narratives might be ideal for Greek tragedy, for example, where the outcome is already fixed, and what matters is how it is reached.
A second solution to the exponential problem in tree narratives is collaboration. Through the World Wide Web (where many interactive narratives already reside) it is technically very easy to open up a tree narrative so that others can add their own sections (The IMRF Choose Your Own Adventure Story is an example). However, this too has its own problems. Constructing a tree narrative in which each path amounted to a small novel and in which every path was complete would be virtually impossible. Rees (1994) points out that even if every single English-speaking person in the world wrote one section of five hundred words, they could only complete a tree with twenty-eight branches; each path would amount to fifteen thousand words. At present it is thought there are about thirty million internet users (Vidal, 1995). A further problem is that of consistency and coherency. Once a tree narrative is sufficiently large, whether or not all its branches are complete, it becomes practically impossible for any one person to have followed every path within it. It is thus conceivable, if not inevitable, that a collaborative writer’s addition to the narrative, while following coherently from the path that writer took through the previous sections, may be completely inconsistent with the story experienced by someone else following a different path but arriving at the same section. This might, however, be avoided either by tight editorial control (which may itself prove difficult) or by ensuring that every branch of the tree was completely independent of all the others.
The open work
An alternative to the tree structure is that of the web, or open structure. Here any of the different sections of a narrative can be accessed at any time, in any order, hence the meaning of the scene is determined not just by the scene itself but by the particular context arising from the previous and subsequent scenes chosen and experienced by the reader. However, it is not hard to envisage the difficulty inherent in creating such a work, and it may not be surprising that there seem to be so few examples of text- or even visually-based open works. It is easier, however, to find examples in other narrative fields, and Umberto Eco (1989) cites the music of Stockhausen, Berio, and Pousseur. One successful visual example might be the Myriorama, which dates from Victorian times and comprises a deck of picture cards which can be arranged horizontally in any order so as to make a single coherent picture.
One example of a textual open work is Mallarmé’s Livre in which the reader would have been able to arrange the pages in any sequence and still experience a coherent narrative. It is perhaps significant that Mallarmé never finished the work.
The problems faced by anyone attempting to author this type of interactive narrative are numerous. One is that it is much harder to deal with the consequences of events – because the story is fluid, the sequence of events flexible, it becomes possible, for example, to experience scenes involving characters who died several scenes previously, thus upsetting any notion the reader may have of the story’s timeframe (anyone who has seen Pulp Fiction will probably have experienced this, although it must be noted that the film does work, but probably only because it is so good in other respects – the unconventional sequence still causes confusion for the first-time viewer). This may not be so much of a problem for the author as for the reader, however. Perhaps the problem is just that we, as readers, are not yet used to the concept of a narrative which has no temporal sequence, or where we can jump back and forwards through time at any moment and with no warning. It may therefore be necessary to adapt the style of the narrative so as to warn, or even to lead the reader to expect, that this may happen. So far this seems to mean the exclusive use of the “wacky, zany, sci-fi comedy” genre in which the reader knows that anything, even the impossible, can happen – for example, The IMRF Choose Your Own Adventure Story or even Douglas Adams’ The Hitchhiker’s Guide To The Galaxy.
Another problem is that of ambiguity or vagueness. If a work is to contain scenes which can be experienced in any order and still be coherent, then it is much harder to make those scenes say anything concrete. The more contexts the scene has to work in, the less specific it can be, the more ambiguous or vague it becomes. But if it becomes too vague, then where is the content of the story? Surely it has been reduced to trivia and as such is unlikely to hold the reader’s interest? Perhaps a solution to this is to make the work open in meaning rather than structure. An example might be Finnegan’s Wake, which Eco (1989) considers to be the the exemplary open work, and which contains many different potential meanings, none of which can be said to be the dominant one. Thus the reader is left to decide how to interpret the text. But one could argue that this is a different kind of openness. Here the work is complete but open to interpretation; the alternative is a work “in movement” as Eco (1989) puts it, where the reader determines the sequence of scenes. Could it be that text is too precise for this purpose and hence does not make a good medium for the open work?
A further issue is that of what happens to closure in the open work. According to Ricoeur (1984), cited in Landow (1992), “to follow a story is to move forward in the midst of contingencies and peripateia under the guidance of an expectation that finds its fulfilment in the `conclusion’ of the story.” The conclusion provides the end point of the story. But can an open work have an end point and hence a conclusion, especially if the work permits the same scenes to be revisited many times in different contexts (eg Michael Joyce’s Afternoon)? Indeed, Joyce’s solution to this problem is to give the reader the responsibility for closure: “When the story no longer progresses, or when it cycles, or when you tire of the paths, the experience of reading it ends.” (Joyce, 1988 cited in Landow, 1992). When the reader stops reading, the story is over. Whether this is successful may, however, be open to question – if a narrative can end at any point, can its plot really be anything more than trivial?
The writer and the reader
A key issue for interactive narrative is that of the role of the author. In the interactive medium, the distinction between reader and author is blurred; the text is more `writerly’, to use Barthes’ terminology. By following links, creating paths, choosing what happens, and what is seen, the reader is creating the narrative, thus taking on part of the role of authorship. The reader, not the author, is now determining points of closure within the narrative. The “function and authority of the author is usurped by the reader” (Cameron, 1993); the role of the narrator becomes redundant. So is this Foucault’s `death of the author’? Probably not – the original author still has to provide the framework, the context, the raw materials, from which the reader creates the narrative.
But doesn’t this necessitate that all readers are good at creating narratives, at telling stories? “Every successful [narrative] form, be it a novel or a feature film or a play or a comic, needs a skilled storyteller to weave together a spell in the mind of the audience, suspend their disbelief and take them on a carefully planned emotional roller coaster through the story” (Cameron, 1993). Perhaps not everyone is a good storyteller, or wants to be. When we read a book, watch a film, listen to a piece of music, we want to be entertained or informed through someone else’s skill; do we really want to have to create the work for ourselves? Do we really want to be authors?
Perhaps, then, interactive narrative is simply not satisfying to read (Rees 1994). Maybe there is an enjoyment to be found in the (comparatively) simple linear structure of the conventional narrative which is just not available from interactive narratives. Perhaps the latter simply does not offer enough to counterbalance the fact that, as Max Whitby, the head of the MultiMedia Corporation, points out (Cameron, 1993), the interactivity gets in the way of the story. In other words, the interactive narrative needs to have something very special to make the reader stay with it `in the face of adversity’. I feel this may be one of the genre’s most significant failings.
However, this does not have to mean the end for interactive narrative. The obvious solution would seem to be to make it more interactive, make the reader more involved. Ultimately, the reader becomes a player, and the narrative a game.
Games and narrative
Conventional narrative refers to the past, it’s temporal referent is `once upon a time’ (Cameron, 1993). By contrast, interactive narrative is in the present, its events are happening now. We are not so much reading about things which have happened as witnessing events in progress. We therefore potentially have the opportunity to interact with these events themselves, to become a protagonist, to be `in’ the story. This is precisely what is offered by the adventure game, or interactive drama as Kelso et al (1992) prefer to call it. Here, you enter “rich, highly interactive worlds, inhabited by dynamic and complex characters, and shaped by aesthetically pleasing stories” (Ibid). The narrative element is provided by the descriptions of the world you have entered, the goings-on within it, and the context in which it is placed – usually a brief conventional narrative outlining the events leading up to the present, and your character’s intended or envisaged role in the plot about to develop. The interaction is present through the player having to solve puzzles or perform actions in order to shape and move through the plot, and the moment of closure arrives when the goal of the adventure is completed. Myst is perhaps the definitive example.
Adventure games are effectively virtual worlds in which the reader becomes player and protagonist, and takes part in an interactive narrative. But is this not true of all computer games, not just those of the `adventure’ genre? And if so, does this mean that all computer games are interactive narratives? Cameron (1993) thinks so. He suggests that a game such as the Hellcats flight simulator “provides a simple narrative framework in which to act…[and] characters to interact with”. Indeed he quotes Borges’ fictional character Herbert Quain to demonstrate that the essential features of games – symmetry, arbitrary rules and tedium – are also shared by interactive narrative.
But isn’t there something counter-intuitive about calling a flight-simulator or a shoot-em-up game an interactive narrative? Are they not, rather, the events about which a story could be told, rather than the narrative itself? In other words, does the narrative reside in the game itself or somewhere outside it, perhaps in the perception of the player? But where, then, does this leave the adventure game? Can we argue that the characteristics of symmetry, arbitrary rules and tedium are not present here? With the possible exception of the last, I think not. Perhaps it is debatable, then, whether the adventure game is interactive narrative, but what we can say is that it fulfils the role which (other) interactive narratives have so far failed to fulfil.
Commercial considerations
The fundamental problem for anyone wishing to market interactive narrative fiction as commercial product is that it doesn’t know which hat it is wearing, or which market it is aimed at. Is it a new form of the novel or film, or is it a type of game? Unfortunately nearly all existing works fall somewhere in the middle and fail on both counts.
As the former they fail because they are not satisfying to read. My feeling is that, when we read a conventional novel, or watch a conventional film, we do so passively – aside from turning the pages, only our imagination is at work (and sometimes not even that). The interactive narrative, however, requires active reading (Landow, 1992), but now the interaction gets in the way of enjoyment, most noticeably (and ironically) when the level of interaction is fairly low – for example, selecting where to go next, choosing between several points of view and so on. Once a sufficiently high level of interaction has been added to overcome this problem, the narrative no longer serves the needs of the passive reader. Now there is too much active participation required – the product has become a game! As games, most interactive narratives also fail, because they do not contain enough interactivity!
It may be that, for the time being, any interactive narrative which is to be successful needs to be aimed either at readers or at players, but not both. If interactive narratives are to appeal to the passive reader (the novel market), an unobtrusive interface needs to be developed, perhaps one which is actually a fundamental part of the narrative rather than simply the means of navigating through it. On the other hand, if they are to appeal to the active player (the games market) then they must be developed as fully fledged adventure games. Perhaps only when we better understand the medium will it be possible successfully to create interactive narratives which appeal to both audiences at once.
The project
The rationale behind our project – One Room, Three Worlds – was to create an environment in which a narrative might be explored interactively. Originally this was to involve looking at how two people’s perceptions and memories of the same events were different – in other words the reader/viewer, by interacting with the characters and objects presented on the screen, would witness an emerging narrative from different points of view. The idea was that each member of the group would take one or more narrative themes and develop them from the point of view of the two characters. However, after the Christmas break, this was changed so that each group member took responsibility for a character (there now being three characters) and developed their narrative. The idea from which the project’s name is derived now emerged – each character was physically present in the same room, but were also in their own personal mental world – the old man remembering the war, the woman having been abandoned, the young girl in her imagination.
Effectively, then, the project was a collaborative one with each of us working independently much of the time. In hindsight, it seems clear that it therefore became subject to the same problem that can occur with collaborative internet-based interactive narrative fiction – namely, lack of coherency. While each individual narrative had a fairly clear theme, there was no obvious or implied link between them. Admittedly the intention was that, given more time, there would have been a greater number of narrative elements, and a better developed plot. Thus the links between the characters would have emerged more clearly. In hindsight it would have been better to have concentrated more on making these links explicit. It may also be seen to highlight one of the problems of open narratives – given that it was possible (indeed the intention) to view most of the different narratives in any order, could an overall coherent narrative really emerge?
In the light of the problems facing open works, perhaps we should not be placing the emphasis on developing an overall narrative in the conventional sense, but on using the narrative elements to create an impression, a mood or feeling, so that the reader/viewer comes away from the computer with the feeling that they have experienced an insight into the lives of the characters, that they have entered their world for a while. Of course, this is not much different to the experience of reading a good, conventional story, except that the notion of plot would have been redefined – no specific event would have occurred but, done well, the reader would be equally satisfied at the end. The emphasis, of course, must be on “done well”. The experience of creating this project suggests to me, however, that this might prove to be a very difficult thing to do. In the same way that hypertext tree fiction needs something special to overcome the fact that the interactivity gets in the way of enjoyment, so the characters in any project like ours must be made to captivate the reader, to make them want to know more about them. Again, `something special’ needs to be incorporated to compensate for the absence of a conventional plot.
An interesting point arises from Barthes’ (1977) statement in Introduction to the Structural Analysis of Narrative that “narrative is present in myth, legend, fable, tale, novella, epic, history, tragedy, drama, comedy, mime, painting, stained glass windows, cinema, comics, news item, conversation.” Given that this project involves presenting narratives visually, and the interface is also visual, what does the interface say? Can a narrative emerge from the interface as well as the characters? I do not think I would cite our project as a good example in this respect, but this may go some way to suggesting a way of designing an interface for interactive narrative which is itself a part of the narrative rather than simply the means of navigating through it.
As a piece of interactive narrative, I do not think our project was particularly successful, mainly because there simply was not enough narrative in it. As an exploration of a possible way in which to present interactive narrative, it shares features which we might expect to see in many future interactive narratives – namely the notion of entering a virtual on-screen world in which the user interacts with objects and characters in order to experience the narrative. I do not think One Room, Three Worlds is a particularly good example, but I do feel this notion of a virtual, visual, environment will be an important medium for interactive narrative fiction.
Conclusion
Clearly anyone attempting to author interactive narrative faces numerous problems which must be overcome if the product is to work successfully. Interactive narrative fiction as a concept undoubtedly has considerable potential, much of which has probably yet to be realised, but if current works are anything to go by, I fear it may be seen as little more than post modern experimentation. It seems to me that there are two ways forward from this position.
One is to decide just what we expect to get from an interactive narrative fiction – is it intended to be used whenever we might otherwise be reading a novel, or when we want to play a game? In other words what category does it fall into – novel or game? Eventually, it will not need to be categorised as either, as it will define its own category, but this is not happening yet. The trouble is that, as with any emerging new technology, we are too bound up in experimenting with the medium, and not paying enough attention to what we are saying. When authors find they have something to say in the medium of interactive narrative, then it will hopefully become a success.
The other way forward is to stick at it (bearing in mind the above). Perhaps the real problem is that as readers we are simply not yet ready for interactive narrative. Perhaps we have not yet become familiar with the ways of thinking required by such work. Further, perhaps writers, too, do not yet know how to write successful interactive narrative for the same reason. If this is true then we can expect interactive narrative fiction to improve. While current works may not be particularly successful, as more people read them, hopefully those who do get to grips with them, those who do get used to the modes of thought required, will produce their own works, attempting to improve on existing ones. In turn, more people will read these, and some will themselves write their own interactive narratives. In other words, a dynamic and dialectical relationship will develop between writers and readers of interactive fiction such that as more readers adapt to the medium, or become `hyper-literate’, more works will be created, and the quality will hopefully improve. As Deemer (1994) states, “for writers and readers alike, hypertext may well define what it means to be literate in the 21st Century.”
As with almost any new medium, the first works created in it are usually based on ideas and experience derived from previous media; later works are usually far more adapted to the medium and hence work more successfully within it. Perhaps interactive narrative fiction has not yet reached this second stage; hopefully it will do so in the very near future.
Bibliography
Books, essays and articles:
Aristotle (384 – 322 B.C.), Poetics, Hackett Publishing Company, London
Barthes, R. (1977), “Introduction to the Structural Analysis of Narrative” in Image, Music, Text, Fontana, London
Bordwell, D., Thompson, K., (1990), Film Art – An Introduction, McGraw-Hill Inc, New York
Cameron, A. (1993), Dissimulations – Illusions of Interactivity
Deemer, C. (1994), What Is Hypertext?
Eco, U. (1989), The Open Work, trans. Anna Cancogni, Harvard University Press, Cambridge, Mass.
Kelso, M., Weyhrauch, P., Bates, J. (1992), Dramatic Presence
Landow, G.P. (1992), Hypertext – the Convergence of Contemporary Critical Theory and Technology, Johns Hopkins University Press,
Baltimore
Rees, G. (1994), Tree Fiction
Vidal, J. (1995), “Bank to the Future”, The Guardian, Jan 28th 1995
Film and video:
Adams, D. (1988), The Complete Hitch-Hiker’s Guide to the Galaxy (video), BBC, London
Tarantino, Q. (1994) (dir.), Pulp Fiction , Miramax Films.
Examples of hypertext fiction on the internet:
Deemer, C. (in progress), Stories from Downtown Anywhere
Firstenberg, A. (in progress), Addventure
Kuwamoto, S. (in progress), The IMRF Choose Your Own Adventure Story
Valimaa, H. (date unknown), Accounting for the Cards
Zacour, J. (1994), The Doomsday Brunette
Commercial Products:
The Voyager Company (1991), The Annotated Alice, (based on the book of the same name by Martin Gardner)
The Voyager Company (1992), Host, (based on the novel by Peter James)
Joyce, M. (1990), Afernoon, Eastgate Systems Inc.