CREDITS: Copyright © 1995 by Michael St. Hippolyte. All rights reserved. Reprinted with permission of the author.
NOTES: This article was originally written with an intended audience of adventure game developers (think Zelda). I have inserted some text to address areas where ARGs are different from these types of adventures. I welcome any commentary or thoughts on the deaddrop.us message forum.
SOURCE: Charted Territory: Nonlinear Narratives in Practice. Reprinted with permission at deaddrop.us
The limitations of nonlinear writing are real and daunting. However, this has not prevented authors from attempting to create interactive stories, and it has not prevented some of the attempts from working effectively. Most of these follow one of three models of interactive storytelling, which we will now examine. Our analysis will include a look at how each of these approaches addresses the fundamental problems described above.
Hypertext refers to a set of viewable information objects which contain embedded, easily navigable links to each other. Although hypertext is more often associated with reference works and pedagogic material than with storytelling, the body of hypertext fiction is growing steadily, driven in no small part by the popularity of the World Wide Web.
The hypertext decision structure is naturally closed; dead ends are rare. It is a perfect decision structure in the sense that Platonic solid is called perfect: every path converges; all paths are explicit; all directions are equal; and every decision is valid. Hypertext narratives have their strongest impact when the links themselves communicate meaning, which occurs when the viewer has traversed enough paths and begins to visualize the narrative structure as a whole. The links can then form a kind of sketch, distilling the story into a single timeless image. On the other hand, if the links fail to carry any meaning of their own, if they are just links, the interactivity is reduced to a procedural convenience for reading the text electronically, and the ability of the viewer to affect the story vanishes.
Regarding the interactive dilemma and the problem of chronology, the author of a hypertext narrative gives up control over the sequence and timing of the story. The rich set of navigational abilities standard in most hypertext environments means that the author cannot typically force the narrative in any particular direction for very long. The number of choices available at one time is usually limited (seven is a typical number), but beyond that the viewer is free to choose which, when, and whether. The narrative challenge of hypertext is akin to writing a story on a deck of playing cards, with a fresh shuffle before each telling. Of course, the viewer's navigational choices are not really random, but for the author they may as well be, at least until hypertext produces its Hitchcock to discover ways to draw a viewer down a story line against his will.
Linear Narratives With Embedded Nonlinear Elements
Another approach to interactive storytelling is to converge the paths so much that the narrative structure is on the whole linear, with isolated nonlinear digressions, usually puzzles or games. Many examples may be found among children's multimedia titles of the popular animated storybook genre. In these examples the story always follows the same path, but the viewer can often click on a character or object and get a sideshow of some kind. If these sideshows do not have any effect on the subsequent story, then they are only locally nonlinear and globally the narrative remains linear. Being globally linear largely avoids the problems of nonlinearity.
It is fair to say that in a story that follows this approach, the interactivity serves as decoration and illustration rather than as an integral part of the narrative. Its most appropriate use is for adapting books and other linear works to multimedia.
By far the most widely used model for interactive storytelling is the maze. The first interactive story built on the maze model was Adventure, a computer game developed on minicomputers in the 1970's. Adventure allowed the viewer to explore a vast system of interconnected underground chambers, moving at will from room to room, collecting clues and other objects for use in solving the central puzzle of the game. In subsequent text adventures, as the genre came to be known, the caves were often replaced by buildings, royal palaces, or Martian bases, but the underlying form remained the same. As graphics of reasonable quality and performance became common, the graphical form of the maze story began to emerge, in which the maze is visually rendered and integrated into the imagery of the story. The vast majority of interactive narratives released on CD-ROM are graphical maze stories.
In contrast to hypertext, the links in a maze story are metaphorical rather than literal. They depend on, and are limited by, an allusion to a physical model of interconnection. This physical model translates the decision tree into a space of some sort (cave, island, haunted house) through which the viewer moves. The paths naturally converge, because they are physical paths in a closed space.
The metaphorical nature of the links is also the maze model's answer to the interactive dilemma. The storyteller depicts the entire decision structure metaphorically (interconnected rooms in a house, for example), works the metaphor into the story as an omnipresent theme (you are always in the house), and uses other elements of the story to limit the viewer's freedom of choice without diminishing the metaphor (you can't get into a particular room because you don't have the key). The viewer ends up perceiving much more freedom of choice than he actually has; this in turn gives the author more control, and more opportunity for expression.
The problem of chronology is not so easily solved, however. Each potential component of the narrative is associated with a physical location, and becomes part of the narrative only when and if the viewer arrives at that location. Some viewers will get to locations faster than others, and not all viewers will visit every location. The player drives the narrative by moving through the space it occupies; thus, whereas traditional stories move through time, maze stories move through space. This substitution of space for time is an obstacle to establishing a chronology. It can leave the viewer with the sense that time has been drained out of the story, that much of the story took place long ago. It can also make the characters in the story seem outside of time, which means less real. This is the price the storyteller must pay to use this model.
One way maze stories solve the chronology problem is by presenting two intertwined stories: the story from long ago that the viewer uncovers and the present story of that uncovering, the former unchanging and the latter created anew with every telling. The unchanging story can deliver the bulk of the narrative information, while the ever-fresh story of discovery can give life to the experience.
From Adventure to Zork to Myst, maze stories have achieved phenomenal popularity. This should not be surprising. The metaphor which lies at the heart of a maze story serves many powerful functions at once: it informs the user how to proceed; it recurs through the narrative, providing a perfect vehicle for establishing a visual or verbal style, setting an emotional tone, or communicating a thematic idea; and, perhaps most significantly, it draws the interactivity into the story itself. The viewer has no choice but to enter the world of the narrative, and on its terms.
The maze story author's primary challenge is to make this world one that the viewer wants to be in and wants to explore. Most often this is accomplished by presenting enticing puzzles for the viewer to solve, leading up to a grand puzzle that becomes the climax of the story and determines its outcome. The most enjoyable maze stories, however, are the ones in which the world of the narrative is intrinsically interesting and worth exploring.