Narrative Production and Interactive Storytelling – Alex Mitchell

Abstract: Many different approaches have been suggested for how stories can be told using interactive media. Forms have emerged which range from the collections of text fragments and links that one finds in hypertext fiction, the adventure-game-like experience of interactive fiction, the story-driven gameplay of Half-Life 2 (Valve Corporation, 2004), through to the one-act-play drama of Façade (Procedural Arts LLC, 2005). All of these forms suggest different underlying models as to what makes a story “interactive”. This paper will explore the ways in which the materiality of the tools used to create an interactive story influence the form of the resulting work, with reference to works created using two different authoring tools: Flash (Adobe Systems Incorporated, 2007), and the NeverWinter Nights Aurora Toolkit (Bioware Corporation, 2002). A comparison of these works will be used to reflect upon the various forms of storytelling that may be possible within interactive media.



For many years, writers and academic researchers have dreamed of interactive storytelling systems such as the Holodeck from Star Trek: The Next Generation (Murray 1997: 15) that will allow the viewer to “step into” the story and become the main character. Much research has been carried out to explore the possibilities for adding interactivity to the experience of stories.[1] But what does it mean to tell an “interactive” story? Although interactivity between the story and the audience is often seen as a new phenomenon, stemming from the advent of “interactive media”, interactive storytelling is something which has been around for millennia. We see an element of interactivity in oral storytelling, where a storyteller conveys a well-known story to an audience, responding to the mood of the audience, and drawing in the audience. When a good storyteller asks “what do you think happens next?” while telling children a bedtime story, for example, she already knows how the story will go, but is deliberately drawing the listeners in, to give them a feeling that they are somehow influencing what happens. We also see interactivity when children play make-believe in the sandbox, making up stories as they go along and constantly re-directing the plot to match the group’s interests. Children playing make-believe involves interactivity: children work together to create a story, with no set rules or guidelines, but instead a socially negotiated story that “feels right” for everyone involved. A similar situation occurs in improvisational comedy or improvisational theatre, such as on the television programme Whose Line Is It Anyway? (BBC Radio 4, 1988; Channel 4, 1988-1998; ABC, 1998-2006). In improvisational comedy, the comedians are responding to suggestions and directions from the audience, and working in a small group to construct a simple narrative to meet the audience’s requirements.

All these examples involve groups of people working together in various ways to construct a narrative that meets the group’s needs. People, unlike computers, are intelligent beings, and as such are able to instinctively construct a narrative on-the-fly in response to interaction. What happens when the text under consideration is not a “live” performance, but instead is an interactive media object such as a hypertext, game, or interactive fiction? What happens when a story is no longer fixed, as in a movie or a novel, and is not being negotiated among a group in a social setting, but instead is conveyed through an interactive media such as a computer, which allows the user(s) to make choices, and changes its responses based on those choices? How can these choices be represented to the user? How can the author express these choices and their outcomes? In computer-based forms of interactive story, unlike those expressed through human-mediated communication, the author is not present during the interaction. Designing an interactive story, therefore, becomes what Salen and Zimmerman call a “second-order design problem” (Salen and Zimmerman 2003: 168): the author/designer needs to design an interactive system that will later respond appropriately to the user as she interacts with the system.

I will discuss the issues involved in authoring an interactive story by looking at the materiality of the tools used to create interactive stories. By materiality I am referring to the materials from which an interactive media piece is constructed, in particular the underlying code which determines the responses of the system to the user, and the ways in which the authoring tools allow the author to think about and manipulate this material. The discussion will take place in the context of two works of interactive storytelling created by my students in my interactive storytelling class at the National University of Singapore. Through this discussion, I will attempt to gain some insight into what it means to tell a story using interactive media, and what tools can help to support the process of creating interactive stories.

What is Interactive Storytelling?

Stories are such an integral part of our lives that we rarely stop and think about what a story actually is. Minimally, narrative consists of three stages: the possibility, the event, and the result (Bal 1985: 189). More complex narratives are decomposed into a sequence of related events, but the general structure holds at any level of complexity. The plot of a narrative generally “is formed from a combination of temporal succession and causality… moving from a stable beginning through complications to another point of equilibrium at the end” (Martin 1986: 81). This traditional form of a plot is derived from Aristotle (Lane 1947: 58), and is conventionally represented as ‘Freytag’s Triangle’, which starts from an exposition, moves through a complication or rising action to a crisis point, and finally to a resolution (Freytag 1968: 105). This movement from beginning to end can also be seen as what Laurel calls the ‘flying wedge’, which consists of “a progression from the possible to the probable to the necessary” (Laurel 1993: 70). Narrative structure can also be seen as consisting of two layers: that of the ‘what’ and the ‘way’, or what Chatman calls the story and the discourse (Chatman 1978: 19). A story may be told in many different ways, and the way in which the story is told may differ across different media.

Interactive storytelling is a form of storytelling that adds some form of ‘interactivity’ to the process of experiencing the story. Ryan defines interactivity as “the computer’s ability to take in voluntary or involuntary user input and to adjust its behaviour accordingly.” (Ryan 2006: 98) The implication here is that the user has some choice – the user may influence the way that the computer responds, as opposed to the computer always behaving in exactly the same, predetermined manner. One of the key issues that arise when adding choice to story is where that choice should be situated within the structure of the narrative, and how to do so without disrupting the narrative structure. As Laurel points out, “introducing new potential… has the capacity to explode the structure of the action” (1993: 73).

Interactivity implies that the reader is no longer passive – that the reader must participate or make an effort of some sort in the process of reading the text. Aarseth introduces the concept of ergodic literature to refer to texts in which “nontrivial effort is required to allow the reader to traverse the text” (Aarseth 1997: 1). By text he means “any object with the primary function to relay verbal information. (Aarseth 1997: 62). He sees information as a string of signs, and distinguishes between strings as they appear to the reader, which he calls scriptons, and strings as they exist in the text, which he calls textons. The mechanism by which the scriptons are generated from the textons as the reader experiences the text he calls the traversal function. Aarseth proposes a set of seven variables that describe how a user traverses a text: dynamics, determinability, transiency, perspective, access, linking, and user functions[2]. These concepts are useful when considering the authoring of interactive stories, as the way in which the interactive story has been constructed essentially determines the textons and traversal function, and thus the non-trivial effort required on the part of the reader.

Ryan breaks down interactivity along 2 axes: external vs. internal, and ontological vs. exploratory. The external vs. internal axis distinguishes the extent to which the user feels that they are in the story. For example, in a game such as Half Life 2 (Valve Corporation, 2004), the player takes on the role of Gordon Freeman, looking through his eyes and directl
y controlling his actions. In a game such as The Sims (Maxis/Electronic Arts, 2000), on the other hand, the player is looking down from above, controlling the actions of several characters in what can be seen as a “god” view of the world. This distinction between external and internal interactivity is somewhat analogous to Aarseth’s concepts of impersonal vs. personal perspectives (Aarseth 1997: 63).

The ontological vs. exploratory axis refers to the extent to which the user’s actions determine the “events” which take place within the system. Exploratory interaction is interaction which allows the user to move around and “explore” a set of pre-determined events, whereas ontological interaction implies that the user’s actions will influence which events take place. This is similar to Aarseth’s user functions. In addition to the interpretive function of the user, which he claims is present in all texts, he suggests the additional concepts of explorative, configurative and textonic functions, although, as Neitzel (2005: 237) observes, Ryan’s ontological interaction collapses configurative and textonic into a single concept, that of ontological interaction.

Based on these two axes, Ryan defines four combinations of interactivity. The first combination is what Ryan calls external-exploratory interaction. In this combination, the user engages in manipulating the ways in which a pre-determined story is presented. This can be seen as what Manovich calls “database stories” (Manovich 2000: 227), where the user is able to explore a set of discourse statements somewhat freely, and over time build up an understanding of the “narrative” through interpretation. These individual discourse statements are what Landow and Delany (following Barthes) call “lexia” (Landow and Delany 2001: 208). Ryan’s second combination is internal-exploratory interactivity. This is very similar to the first combination, in the sense that the user’s actions have no impact on the outcome of the story. However, the user’s position is presented as an inhabitant of the world of the story. This is the least common form of interactivity, as the purely exploratory nature of these types of systems impose restrictions on the agency of the user.

Ryan’s next combination is that of external-ontological interactivity, in which the user takes an external, god-like view of the world, making decisions about how events proceed but not identifying with any of the characters within the storyworld. This is the form of interactivity that can be seen in simulation games such as The Sim. Finally, Ryan identifies internal-ontological interactivity, in which the user takes on the role of a character within the time and space of the story’s virtual world. In these systems, a story is generated from a set of rules as the user interacts with the system. Here, the idea is to balance between the author’s concept of the story and the user’s desire to take action. Allow the author’s vision to dominate, and the experience becomes exploratory rather than ontological. Allow the user complete freedom, and the narrative risks becoming ill-formed, as the relationship of the various events within the narrative are not clear to the user.

A good example of a work that attempts to balance this tension between author control and user control is Façade (Procedural Arts LLC, 2005). Façade is an interactive drama which models the experience of the main character (the user) visiting a couple, Trip and Grace, for dinner. During the course of the evening, the main character is drawn into the drama revolving around the breakdown of Trip and Grace’s marriage. In Façade, the player experiences the story through a first-person-perspective interface, looking through the eyes of the main character. The user makes choices by moving around a 3D space using the mouse, and talking with Trip and Grace by typing in dialogue through a natural-language interface. The underlying system interprets these speech-acts through a process of matching the utterances against the current state of the system and then determining the appropriate responses from the two characters. The choices the user makes, through these utterances, can alter the course of the story, leading to seven different possible endings. Façade focuses on providing the player with agency. A term originally applied to interactive storytelling by Murray, agency is “the feeling of empowerment that comes from being able to take actions in the world whose effects relate to the player’s intention.” (Mateas 2004: 21) According to Mateas, agency is the primary component to be addressed in interactive storytelling. Façade embodies the notion of a balance between authorial intention, in the form of formal plot structure and material constraints, and user agency. The system works to balance the actions of the user, which often consist of attempts to disrupt the narrative, with the overall trajectory of the story (Mateas and Stern 2005: 6).

Narrative Production

How does an author go about creating a story? This is the process of narrative production. According to Talib, narrative production can be studied from a systems perspective, where the environment that gives rise to, or has the potentiality to give rise to, a narrative, is the focus of study:

The environment gives rise to a set of possible choices… [A]mong the choices are, whether to resort to narrative rather than other modes of dis­course, or possibly, to write one’s story rather than to tell it orally, etc.  The initial choices can be described as paradigmatic.  Once these choices have been made, the narrative producer tries to arrive at the tenta­tive or even sketchy overall plan of the story itself, which involves some kind of syntagmatic design.

During the act of telling or writing, the narrative producer can be described as engaged in choosing a series of paradigmatic and syntagmatic options, as each set of vertical options is chosen with the overall horizontal narrative ‘structure’ in mind.  The points in the story where further paradigmatic options are made can be des­cribed as reference points where the storyteller or writer may return to when revising or re-planning his or her story: these may also be points that allow the narrative pro­du­cer to back­track when necessary. (Talib 2007:

When writing a story, this planning and re-planning involves imagining and then piecing together elements of the narrative, such as character, scenarios, and events (the story); and deciding on which events to explicitly show, and the order of presentation of the events (the discourse). When developing an interactive media work, the analogous process may involve deciding on the rules which will define and constrain what the user can do, how the user’s actions are interpreted by the system, how the user can move around the system, and how these actions affect both the underlying story and its presentation.

Working with interactive media can be quite different than working with traditional media. The process of creating an interactive work involves the setting up of possibilities, constraints and pathways for user, balancing of choice and control, deciding where the user can and cannot go, what order they can do things, what may happen that the designer/author hasn’t thought of, and so on. The environment of production includes not just the cultural and social setting, but also the media that will be used to convey the story, and the tools that are used to construct the interactive environment. The author’s choice of media and tools can be considered paradigmatic choices, which are then followed by, and influ
ence, the syntagmatic choices made during the design and implementation of the interactive story.

The Materiality of Interactive Media

When I refer to the materiality of interactive media, and the materiality of the tools used to create interactive works, what I mean is the material from which the work is constructed, and the ways in which the tools allow the writer to inscribe their ideas and concepts into that material, in this case interactive media. Here it is not the screen, keyboard, mouse, storage devices, CPU, and other physical devices that are being considered (what Kirschenbaum (forthcoming) calls forensic materiality), although these physical materials are an important part of the experience of the text. What I am considering is the underlying code (Kirschenbaum’s formal materiality) from which the work is constructed, the abstractions with which this underlying code is represented by the authoring tools, and the paradigms for thinking about authoring which these representations embody. Using Aarseth’s terminology, the tools that the author is using determine the ways in which the author can structure the underlying textons and traversal function of the text. This structure will in turn generate the scriptons encountered by the user/reader as she traverses the text.

As discussed by Voithofer, working with interactive media involves the use of “a common set of design resources, constrained by the interface metaphors and technical discourses of the production software… The computer presents epistemological constraints as it offers opportunities for representation.” (Voithofer 2005: 9) The form of the tools or production software has specific affordances (Norman 1998:13) which make it easier to express ideas and concepts in some ways, and harder to express them in others. As Hayles explains:

The electronic author [will] consider what behaviors and animations should attach to the words, in what font and color they should appear, on what background and over (or under) what layers, to what other texts or structures they should be linked, and so forth. In all these activities, the hardware and software are active partners, facilitating and resisting, enabling and limiting, enacting and subverting. The labor needed to program these effects must be seen as intrinsic to the work of creation. Like the creator of an artist’s book who manipulates an Exacto knife to make delicate cutouts in heavy white Italia paper and painstakingly sews the pages together, the writer of an electronic text is intensely aware of the entwining of intellectual, physical, and technological labor that creates the text as a material object. (Hayles 2004:18)

Thus, the syntagmatic choices that the author makes during narrative production will be influenced by the ways in which the tools used serve to structure her thoughts.


To explore how the tools used to construct interactive stories influence the form of the resulting work, I will discuss several interactive stories created in my interactive storytelling class at the National University of Singapore. This class, UAR2205 Narrative and Play in Interactive Media[3], exposes students to the issues surrounding telling stories in interactive media. This class is part of the University Scholars Programme, which takes in students from across the university, from a variety of academic backgrounds. Basic concepts of interactivity, narrative, and storytelling are covered, and students are asked to create two interactive stories.

The first assignment asks the students to recreate an interactive version of an existing story, such as a fairy tale, movie, or short story, using any interactive medium. Students were given a brief introduction to HTML and Flash (Adobe Systems Incorporated, 2007). Students worked in groups of 5 or 6. Most groups chose Flash, with one using a mix of live action and Flash, and one using live action. I will be discussing one of the Flash works. The second assignment was to create a short story in the NeverWinter Nights Aurora Toolkit (Bioware Corporation, 2002). For this assignment, the students were told to think about creating a “short story” through interaction and play, with ‘short’ defined as approximately 15 minutes of interaction to complete the story. As will be discussed below, the choice of tools used influenced the way in which the students thought about the issue of “interactivity” within their works, and the way in which they attempted to incorporate interaction into their stories.

The Others: Flash-based Interpretation of a Movie

One of the most interesting works created by the students using Flash was a retelling of the movie The Others (Alejandro Amenábar , 2001). This work was created by Spencer Hsu, Jesse Gozali, Lionel Sng, Liana Tang, Shamantha Yan and Shane Yan. The original movie is a complex, multi-layered narrative that explores issues of memory and perception. The gradual uncovering of the story, which requires the viewer to actively construct an understanding of the connections between events, is an interesting starting point for creating an interactive story. The fact that much of the story as presented through the film relies on the perception of the characters within the story, with a revelation and reversal at the end, much like other contemporary horror films (The Sixth Sense (M. Night Shyamalan, 1999) is the most obvious comparison), suggests the possibility of converting the experience of watching the movie into an ‘interactive’ experience.

The students’ reworking of The Others was implemented in Flash. Traditionally, Flash has been targeted at interactive media designers who want to create responsive, rich media applications for the web. Reminiscent of non-linear video editing tools such as Premier (Adobe Systems Incorporated, 2007) and Final Cut Pro (Apple Inc., 2007), Flash’s user interface has a stage and timeline metaphor. The stage represents what the user can see at any given time, whereas the timeline represents how the stage changes over time. Each instant in time, as represented on the timeline, is called a “frame”. By default, a Flash movie plays from the first frame to the last, showing the content of the stage in each frame, at a given frame rate. Program code, written in a language called Actionscript, allows the designer to control the ways in which the Flash movie is played, including stopping the timeline, waiting for user input, and jumping to different frames in the timeline. It is also possible, in the most recent version of Flash, to write complex code that completely determines the interaction within a Flash movie. However, most non-programmers tend to make use of the stage-and-timeline authoring environment, rather than writing ‘raw’ code. This stage-and-timeline metaphor, which encourages the creation of a set of rich media fragments linked in a non-linear fashion by user actions, forms the underlying material with which the author is working when creating an interactive story in Flash.

The interactive version of The Others as created by the students involves the user exploring fragments of the underlying story, encountering different segments of the movie, and gradually uncovering the overall narrative. The user moves through nodes in the story, with a ‘map’ provided to help the user to visualize the nodes and links. The work is what Ryan would call external-exploratory interactivity. The user is controlling, within the bounds set by the author, the order in which they encounter the various discourse statements, in the form of hypermedia fragments, which call to mind Landow and Delany’s concept of lexia as used in hypertext fiction (Landow and Delany: 211). As with a text-based hypertext, the work consists of a nodal, database-like arrangement. At the level of the individual lexia, the user encounters linear segments, consisting of still images, audio clips and video fragments.

There are two modes to the interface: a ‘scene’ from the movie, and a ‘map’. In the ‘scene’ interface, the user sees an image from the movie, with invisible ‘hot-spots’ which show a ‘hand’ cursor when moused over. Clicking on a hot-spot will lead the user to another segment. These segments consist of images, audio clips and video clips extracted from the original move. The segments are not linked in a linear sequence, but instead are organized so as to allow the user to gradually uncover more and more details of the underlying story. This gradual revelation of ‘what happened’, and the use of ‘hidden’ links, is similar to the use of text fragments and hidden links in, for example, Joyce’s Afternoon, a story (Eastgate Systems, 1990) created using the Storyspace (Eastgate Systems, 1987) hypertext authoring tool. Storyspace allows the author to create text fragments, which can then be connected by unidirectional or bidirectional links. These links can be made visible to the reader, or can be hidden, only to be triggered on a mouse click. The interactive version of The Others uses a similar technique of leading the reader through a series of seemingly disconnected discourse statements, which the reader will, through the process of interpretation, gradually piece together into a coherent narrative.

The system also provides a ‘map’ interface, which shows the nodes and the links between them in a visual interface. Nodes which have been visited are clear, whereas nodes which have not yet been visited are partially obscured, with their location in the network of nodes visible, but their labels ‘greyed out’. The gradual revelation of the map mirrors the gradual recovery of the memory of the main character in The Others. This use of a visual map is similar to the visual layout of lexia in Jackson’s Patchwork Girl (Eastgate Systems, 1995), which uses both the map view provided by Storyspace, and a graphical image of the ‘patchwork girl’ which links to various segments of the story. Storyspace allows for various views of the story space, such as a map, chart, outline and treemap, which can theoretically allow the reader to move through the lexia independent of the links. It is up to the author, however, to enable or disable these different views. Some works, such as Afternoon, a story, deliberately disable these views, whereas others, such as Patchwork Girl, make use of them as alternative ways of experiencing the narrative.

One interesting feature of the interactive work is the use of ‘locked’ nodes. When the user attempts to follow certain links, they are denied access, with the message “I haven’t remembered enough…” These nodes represent pivotal events which reveal much of the underlying ‘truth’ of the story. These nodes are unlocked once the user has viewed other nodes that foreshadow the events of the locked node. This use of locked nodes is much like the ‘guard fields’ in StorySpace, which allow the author to specify that a link can only be followed after either another link has been followed, or a specific lexia has been viewed. This gives the author some control over the order in which the reader will be exposed to the discourse statements.

The structure of the interactive version of The Others closely follows the timeline-based form made available by Flash. The basic stage-and-timeline metaphor of the Flash authoring environment encourages the creation of linear sequences of media-rich narrative fragments. Flash also makes it easy to respond to user input and jump non-linearly between frames. This directly encourages the creation of a work such as The Others, where a set of linear multimedia segments are linked together in a network of nodes, which are navigated by the user through links activated by mouse clicks. The ease with which video, audio and images can be incorporated into a Flash movie make the deconstruction of a linear narrative into a database-like non-linear web of lexia a natural approach to take when working with this material. This, combined with the layered and potentially non-linear nature of the source text, strongly influenced the final form of the interactive story created by my students. The features of the underlying Flash system, and the way in which the Flash authoring environment makes it easiest for the author to create a specific form of interactive story, represent the influence of the materiality of the tool on the process of narrative production.

The Marketplace: A Short Story in NeverWinter Nights

This work, titled “The Marketplace”, was constructed using Bioware’s NeverWinter Nights Aurora Toolkit. The Aurora Toolkit is a tool for creating new modules in NeverWinter Nights (Bioware Corporation, 2002), a computer-based fantasy role-playing game. The toolkit allows the author to construct areas (spaces), and then populate these spaces with buildings, props, non-player characters, and objects that the player can pick up, use, and interact with. The author can also can link spaces together, add music and other effects, and set event ‘triggers’ within the space. The toolkit also lets the author build ‘dialogue trees’ for non-player characters, which lets the player progress through a branching dialogue. Responses can be altered using variables, which can be used to store the outcome of previous actions. Responses can also be based on vari
ous other conditions, such as a player character’s alignment (good/neutral/evil and lawful/neutral/chaotic) or the objects the character is carrying. In addition, the underlying game, NeverWinter Nights, incorporates a ‘combat system’ derived from the d20 rules developed by Wizards of the Coast, originally used in the table-top role-playing game Dungeons and Dragons[4] (Tactical Studies Rules, 1974, 1977, 1989; Wizards of the Coast, 2000, 2003). This combat system provides a set of rules with which the game system can resolve combat between players and non-player characters. This additional layer of rules, which works alongside the dialogue and interaction with objects, adds additional complexity to the system when it comes to designing an interactive story. The Aurora toolkit allows authors to construct stories which revolve around character and space, and which adapt to the player’s actions. As we will see, the material provided by the tool, which focuses very much on moving through spaces and interacting with characters, affords specific forms of storytelling. In addition, the combat system introduces a gameplay element to the tool which may encourage authors (or players) to take a more game-like approach to designing interactive pieces.

“The Marketplace” is a short interactive story created by Chong Weien, Lucas Ho, Mikail Lo, Aileen Ng, and Pranav Piyush. The story starts with the main character arriving at the main gates of an unnamed town, at which point the character is stopped by the town guards and told that the town is under quarantine, so that the main character can enter the town, but can’t leave. However, as the main character has a delivery to make, it is up to the player to figure out how to leave. There are people scattered around the marketplace with whom the main character can interact, in an attempt to convince them to provide help to get out of the town. The marketplace is a very constrained space, which mirrors the narrative form. There are limited options, and the player needs to work within those options to find her way to the solution of the puzzle. Interaction consists mostly of walking from person to person, getting quests and figuring out how to solve them, with some combat.

This interactive piece can be viewed as a puzzle or a riddle, much like interactive fiction (Montfort 2005: 4). Completing the puzzle requires trying various different combinations of interactions with the townspeople to see which of them will enable you to get past the guards and leave the town. The structure is what Ryan would call level 3 interaction. It is ontological in the sense that there are a variety of paths that can be taken to the end of the story, the outcome of which depends on the actions of the player. It is also exploratory, as much of the game involves the player moving through space to uncover information from either the objects within the world, or through dialogue with non-player characters. Ways of escaping from the town include working for the underworld to get enough money to ‘buy’ papers from Sir Olric, helping Sir Olric’s daughter to get on his good side, or doing odd jobs so that you can afford to ‘buy’ papers. Thus, there are several possible ways to solve the puzzle/complete the story. The player can start down each of 3 different story arcs, each of which consists of 4 mini-quests. Once the third mini-quest is started, then the other story arcs are no longer available. This gives the player the illusion of choice, but actually narrows down the space of possibilities, in a manner similar to Laurel’s notion of the ‘flying wedge’. This illusion of choice is echoed by the spatial layout, which allows player to wander around, thus giving an illusion of freedom of movement. However, it is just an illusion, as many doors are just facades, and can’t be entered. One problem with this approach is a there is a tendency for the player to wander around such a space with little direction or motivation. The Marketplace attempts to deal with this by putting ‘arrows’ in the space, which help to direct the player, but tend to break the feeling of immersion and suspension of disbelief.

The introduction of combat as part of the storytelling process leads to a work that is both exploratory and ontological. It also gives the story an element of what Juul (2006: 67) would call both an emergent and progressive structure. It adopts what Jenkins (2004: 126) calls embedded narratives – the story emerges through exploration of the space – and also an element of emergent narrative, as the player uses the properties of the game system to influence the outcome of combat with NPCs.

The use of the Aurora Toolkit to create The Marketplace directly influenced the form of interactivity that the authors of the work embedded into their story. The Aurora Toolkit allows for the creation of virtual spaces, which are populated by characters. These characters can be given sets of dialogue which the player can encounter through interaction, and can also contain triggers for other events within the world. In addition, the underlying game engine contains a combat system. These material characteristics of the tool determine the textons and traversal function of the works created in the tool. Much of the player’s experience of the story in The Marketplace consists of moving around a series of spaces, which are populated by characters waiting to be spoken to. Solving the ‘puzzle’ of the story involves a combination of saying the correct dialogue to the correct characters in the correct order, and killing certain characters. This is a direct reflection of the materials provided by the authoring tools, which afford the construction of spaces, creation of dialogue trees, and in-game combat.

Implications and Conclusion

As discussed above, the choice of tools for interactive storytelling influences the types of stories told, and the form these stories take. In particular, the affordances of a particular tool influence the ways in which the work incorporates interactivity. Different tools provide different amounts of expressive power to the author in terms of determining how players will be able to make choice within the interactive system. A tool such as Flash, which encourages the use of timeline-based multimedia with conditionals and user input to jump to different frames, affords interactive stories that resemble rich-media versions of the hypertext fictions created using tools such as StorySpace. The resulting work tends to incorporate external-exploratory interactivity, with rich multimedia content but little or no user agency. A game-like tool such as the Aurora Toolkit for NeverWinter Nights encourages a combination of embedded narrative through its use of virtual spaces and character-based dialogue trees, and more game-like emergent behaviour due to the inclusion of a d20-based combat system. The interactivity tends to be internal, focusing on the actions of a specific character, and combines exploratory and ontological interaction as a result of the virtual space and the combat system.

The fact that the tools discussed above tend to encourage specific ways of thinking about interactive storytelling suggests that a more flexible authoring system, which gives the author the ability to mold the tool to suit her needs, may lead to new ways of thinking about interactivity in storytelling, as the author will not necessarily be constrained by the affordances of the authoring tool. A general-purpose programming tool such as C++ allows this flexibility. However, the high degree of expertise required to build an interactive story, even given a game engine such as the Unreal Engine[5] (Epic Games, 2001-2007) or the Source engine[6] (Valve Corporation, 2004), makes it difficult to see this as a path for
potential authors of interactive stories to take. Other tools, such as the RenPy (PyTom, 2004-2007) visual novel engine[7], and the Inform (Graham Nelson, 1993-2007) interactive fiction authoring tool[8], follow specific paradigms of storytelling. These tools work very well for the tasks they were designed for. However, given my observations of how the paradigm of the tool influences authors’ choices when constructing an interactive story, it is safe to assume that these tools will similarly constrain/afford specific forms of interactive storytelling.

When telling stories in a social setting, such as children playing make-believe, the rules for deciding the outcome of the story are malleable, and can be negotiated within the group. However, when an author is working with an interactive media authoring tool which affords specific ways of telling stories, the assumptions about what an interactive story is have already been embedded into the system. This suggests that a tool which is itself configurable and malleable, and can be moulded to fit, on the fly, the needs of the storyteller, rather than encouraging the storyteller to fit one specific model of interactivity, will allow authors to explore new, innovative ways of incorporating interactivity into the process of storytelling.



Aarseth, Espen. 1997. Cybertext: Perspectives on Ergodic Literature. John Hopkins University Press.

Bal, Mieke. 1985. Narratology. Toronto: University of Toronto Press.

Chatman, Seymour. 1978. Story and Discourse. Ithaca, New York: Cornell University Press.

Cooper, Lane, trans. 1947. Aristotle on the Art of Poetry. Ithaca, New York: Cornell University Press.

Freytag, Gustav. 1968. Freytag’s Technique of Drama. Trans. Elias J. MacEwan. Ayer Publishing.

Hayles, N. Katherine. 2004. “Print is Flat, Code is Deep: The Importance of Media-Specific Analysis.” Poetics Today 25 (1): 67-90.

Jackson, Shelly. 1995. Patchwork Girl. Eastgate Systems.

Jenkins, Henry. 2004. "Game Design as Narrative Architecture." In First Person: New Media as Story, Performance, and Game, edited by Noah Wardrip-Fruin and Pat Harrigan, 118-30. Cambridge, Mass.: MIT Press.

Joyce, Michael. 1990. Afternoon, a story. Eastgate Systems.

Juul, Jesper. 2006. Half-Real: Video Games between Real Rules and Fictional Worlds. Cambridge, Mass.: MIT Press.

Kirschenbaum, Matthew. Forthcoming. Mechanisms: New Media and Forensic Textuality. Cambridge, Mass.: MIT Press.

Landow, George and Delany, Paul. 2001. “Hypertext, Hypermedia and Literary Studies: The State of the Art.” In Multimedia: From Wagner to Virtual Reality, edited by Randall Packer and Ken Jordan, 206-216. London: W. W. Norton and Company.

Laurel, Brenda. 1986. Towards the Design of a Computer-Based Interactive Fantasy System. Ph.D. Diss., The Ohio State University.

Laurel, Brenda. 1993. Computers as Theatre. Boston: Addison-Wesley.

Manovich, Lev. 2000. The Language of New Media. Cambridge, Mass.: MIT Press.

Martin, Wallace. 1986. Recent Theories of Narrative. Ithaca, New York: Cornell University Press.

Mateas, Michael. 2004. “A Preliminary Poetics for Interactive Drama and Games.” In First Person: New Media as Story, Performance, and Game, edited by Noah Wardrip-Fruin and Pat Harrigan, 19-33. Cambridge, Mass.: MIT Press.

Mateas, Michael and Stern, Andrew. 2005. “Procedural Authorship: A Case-Study Of the Interactive Drama Façade.” In Proceedings of Digital Arts and Culture (DAC) 2005. Copenhagen.

Montfort, Nick. 2005. Twisty Little Passages: An Approach to Interactive Fiction Cambridge, Mass.: MIT Press.

Murray, Janet. 1997. Hamlet on the Holodeck. New York: Free Press.

Neitzel, Britta. 2005. “Narrativity in Computer Games”, in Handbook of Computer Game Studies, edited by Joost Raessens and Jeffrey Goldstein, 227-245. Cambridge, Mass.: MIT Press.

Norman, Don. 1998. The Design of Everyday Things. Cambridge, Mass.: MIT Press.

Ryan, Marie-Laure. 2006. Avatars of Story. Minneapolis: University of Minnesota Press.

Salen, Katie and Zimmerman, Eric. 2003. Rules of Play: Game Design Fundamentals. Cambridge, Mass. MIT Press.

Talib, Ismail. 2007. Narrative Theory,, accessed 30 June 2007.

Voithofer, Rick. 2005. “Designing New Media Education Research: The Materiality of Data, Representation, and Dissemination” Educational Researcher 34 (9): 9.


Media Works

Afternoon, a story. 1990. Michael Joyce. Eastgate Systems.

Dungeons and Dragons. 1974, 1977, 1989; 2000, 2003. Tactical Studies Rules; Wizards of the Coast.

Façade. 2005. Procedural Arts LLC.

Half-Life 2. 2004. Valve Corporation.

NeverWinter Nights. 2002. Bioware Corporation.

The Others. 2001. Alejandro Amenábar.

Patchwork Girl. 1995. Shelley Jackson. Eastgate Systems.

The Sims. 2000. Maxis/Electronic Arts.

The Sixth Sense. 1999. M. Night Shyamalan.

Whose Line Is It Anyway? 1988; 1988-1998; 1998-2006. BBC Radio 4; Channel 4; ABC.


Authoring Tools

Flash. 2007. Adobe Systems Incorporated.

NeverWinter Nights Aurora Toolkit. 2002. Bioware Corporation.

Storyspace. 1987. Eastgate Systems.

Premier. 2007. Adobe Systems Incorporated.

Final Cut Pro. 2007. Apple Inc.

Unreal Engine. 2007.  Epic Games.

Source engine. 2004. Valve Corporation.

RenPy. 2007. PyTom.

Inform. 2007. Graham Nelson.



[1] See, for example, Brenda Laurel, “Towards the Design of a Computer-Based Interactive Fantasy System” (1986), and Michael Mateas, "A Preliminary Poetics for Interactive Drama and Games" (2004).

[2] Dynamics refer to whether the number of and contents of the scriptons and textons in the text are static or dynamic. Determinability concerns whether the result of the traversal function is always the same, or whether there is some amount of randomization. Transiency refers to whether the text changes over time, in which case it is transient, or if it only changes based on user actions, in which case it is intransient. The perspective of the text is personal if the text requires the user to play a strategic role as a character within the world of the text, and is impersonal otherwise. Access can be random, if the user can access all scriptons at any time, or controlled otherwise. Linking refers to whether there are explicit, conditional, or no links between scriptons. Finally, user functions consist of interpretative, explorative, configurative, and textontic user functions. The interpretative user function, Aarseth claims, is present in all texts. The explorative function requires the user to decide which path to take through the text, the configurative function involves the user in part choosing or creating scriptons, and the textonic function which allows the user to permanently add textons or traversal functions to the text.

[3] See for details on my class.

[4] See for details about Dungeons and Dragons.

[5] See for information on the Unreal Engine.

[6] See for information about the Source engine.

[7] See to find out more about the RenPy visual novel engine.

[8] See for more details about Inform.

Author Bio

Alex Mitchell teaches interactive media in the Communications and New Media Programme (Faculty of Arts and Social Sciences) at the National University of Singapore, Before joining NUS, he was a lecturer at the School of Design, Nanyang Polytechnic, where he taught and developed projects in interactive media and games. Alex has worked as an interaction designer at IDEO, London, and at Kent Ridge Digital Labs,Singapore. He has an M.Sc. in Computer Science (Human-Computer Interaction) from the University of Toronto. His work has been shown at SIGGRAPH’98, at the Science Museum in London, at Graphite 2004 at Nanyang Technological University, Singapore, and as part of the Creative Curating Lab at the Singapore Art Show 2005.

Contact Email: