As a rule, works of interactive fiction, or IF, seek a narrative of some sort and must therefore have a setting for the narrative to happen in. The preeminent method for organizing this setting is to employ a metaphor in which the setting is comprised of one or more rooms. These rooms are typically connected to one another and can be navigated via compass directions as well as a few variations, with up and down or inside and out being among the most common.

The room metaphor dates back to the earliest text adventure games and MUDs. Many games in those days took place in settings for which the room metaphor was natural: Roberta Williams chose a Victorian manor for her early game Mystery House, and Crowther's Adventure was inspired by the Mammoth cave system beneath Kentucky. Each room could be given its own description or scenery, and the game designer could be afforded with a structure that lent itself well to modularity. His magnum opus could be subdivided into manageable pieces that mapped directly to the domain of the game itself, and the ramifications of changes in one piece could be easily understood in the context of another. There are games of increasingly large scope that are similarly organized to this day, and future games are likely to continue the practice. The technique of room as metaphor has many successes to its credit, and for any would-be IF author planning out his next piece it is a tool well worth considering.

Even in cases where the metaphor of interconnected rooms is not a very apt description for the game itself, the technique has still proven useful. Consider a city divided into streets or even parts of streets, or complicated layouts that contain one room inside another. For instance, a theater (which is a room) might contain a stage (another room) and beside it a pit for the orchestra (yet another room). Here we have a setting that is, strictly speaking, a single room, and yet if the three perspectives from the theater, stage, and orchestral pit are all crucial to the story, we may decide as a designer to model the setting as three separate rooms. Indeed, there may be even more rooms than that, if a moment served standing in the aisle or huddling crouched beneath a spectator's seat was to offer some valuable perspective as well.

However, with such a potent tool as this we run the risk of having it color our game in unwarranted ways. We may even find ourselves putting on blinders with respect to what a work of interactive fiction is capable of, and may miss out on possibilities that a world full of rooms might tend to obfuscate. The purpose of this discussion, then, is to more carefully inspect the metaphor of a room in interactive fiction, determine how and why this affects the very nature of a game, and what some alternatives to the metaphorical room approach might be.

Tangibility

A corollary to the fact of rooms being tangible metaphors is that they tend to be observable, and as such the faculties of an in-game observer are needed for the player, in turn, to be able to observe them. This implies that a protagonist or in-game personality needs to possess a particular perspective in order for the player to know about that room. Since we are talking about physical spaces, it follows that the perspective is one of location. The idea of a protagonist being located at a specific place at any point of time is one that is central to interactive fiction, and one that falls very naturally out of the idea of an IF piece consisting of rooms. Indeed, for those of us accustomed to this structure it is difficult to think of it in almost any other way.

However, if we consider alternatives that are intangible, we begin to speculate about how few of IF's potential surfaces we have actually scratched. One possibility that would seem altogether obvious from a novelist's perspective (though not, perhaps, from a game designer's) is the idea of dividing a work into chapters. Each chapter, of course, is free to acknowledge the existence of many rooms, or even no rooms at all. And in contrast to a novel we are free to present these chapters in any order we like, even as many times as we like. But the important difference here is that we have replaced the edifice of tangibility with one of structure, and can use this to view our work in a different way as we compose it. This leads to the potential for other conventions that follow naturally from the chapter metaphor, for instance the use of chapter titles, or footnotes, or tables of contents as a sort of mapping mechanism. All the while we can address the limitations of such constructs by keeping in mind that interactive fiction is, above all, interactive, and that our table of contents or footnotes may be much less static than we are accustomed to.

And if we move from the novelist's chair to the playwright's, we might consider the notion of scene as a guiding structure. This metaphor, with respect to rooms, may be thought of as the inverse of a chapter, for while a chapter will typically involve multiple locations, a room may host a series of scenes. If anything, the idea of a scene as a unit of divisibility stresses the location of the action more than ever, and indeed, it becomes easy to envision a work divided into many scenes, each of which share a single room.

In fact there is an encouraging movement in IF for the development of single room games. There are many benefits to this approach, foremost of which is a tendency to limit the work's scope, thereby increasing the likelihood it can be thoroughly completed and tested. But a more subtle benefit is that it veers from the path of room-based gaming, and allows us to explore some rarely traveled ground. Andrew Plotkin's Shade, for instance, achieves a claustrophobic effect by confining its action to a single cramped apartment. A designer mired in the context of multiply-connected rooms might have been unaware that such effects were even possible in interactive fiction.

Spatial vs. Temporal

IF's focus on rooms necessarily brings us to considering the work in terms of spatial units. It doesn't take a great leap in intuition to go from rooms to other related concepts such as the scenery and objects that inhabit these rooms, and to the idea of maps that join all these rooms together. In fact a large part of the IF tradition is one of the explorer protagonist, and by focusing on spatial metaphors we are encouraged to ask questions that facilitate this tradition. How often have we wondered what lies beyond the next door, or what creatures might lurk in the pit that yawns before us? Using rooms as our unit of structure logically results in a containment hierarchy, one that describes a tree of objects and actors organized according to their relative location in the game world.

An interesting side effect of this arrangement concerns the nature of puzzles, which are frequent inhabitants in the world of interactive fiction. It should come as no surprise that many IF puzzles will be somehow spatial in both their nature and consequent solution, seeing as how they are housed in fortresses that are characterized by their spaces.

But what if you want a world full of rooms that move, shift, or come from nowhere, only to later disappear altogether? Such a game, of course, would still be possible with today's room-based systems, although a wise way to implement them might not be particularly obvious. The Adventurer's Consumer Guide by Oyvind Thorsby is one such example of this: after a wonderfully strange interlude with the Queen of Finland in one of the game's middle parts, which may or may not take place outside of time and space for its own sake, the player who goes on to revisit some of the game's earlier rooms will find them chilling reminders of what they once were. And if the idea of rooms that move is central to your game's theme, the room metaphor may become inconvenient or even unworkable. This latter scenario is probably unlikely; after all, the room is only a tool to help guide our thinking, and can be surmounted or ignored when the situation calls for it. However, it is still valuable to consider alternative metaphors, even if the set of all possible works of IF can be crafted with our existing ones. Our goal in thinking about these things, after all, is not so much to make new things possible as to make them obvious, and to explore the structures that enable us to achieve that.

The obvious antipode of a spatial division is a temporal one. Imagine a time-travel work divided into years, or a courtroom drama divided into the stages of a confession. An account of a soldier's rise to power could be subdivided into the ranks he achieves throughout his career. The documented history of a town could be subdivided according to its epochs or historical periods.

None of these examples require us to dispose of rooms completely, just as room-based games are not incapable of describing a story that takes place over a long period of time. The difference, though, is one of degrees, and of whether time is the primary structural factor instead of a supporting one. Just the act of looking at a game's world through a temporal lens may affect its nature, even if it is actually implemented as a computer program using a spatial one.

Crowdedness

Rooms by their very nature are crowded places. They can be observed all at once, they are typically parts of a logical whole (if not, we might do better to call them houses), and a character's actions in a room are typically there for the entire room to witness. Rooms can be intimate and are prone to breeding familiarity. We live our lives in rooms and we populate them full of things to reflect our goals, our interests, and our fantasies.

But what if our story is not an intimate one? What if it isn't a story at all? What if we are seeking to document a vast and impersonal place, where rooms prove too fine-grained? Surely there are ways to get around these problems when restricted to a room-based structure, but why should there be a restriction at all?

If we are to enlarge the room into something else, at least two possibilities come to mind. The first is the notion of a level, which is a common metaphor in the adventure and RPG genres. With a level we still respect the spatial division, but we enlarge it to allow for a sense of strategy and tactics. How many times could we have been liberated, like Photopia's angel in the labyrinth, if we had seen the level before the room? And surely a first-person shooter would be much less compelling if it took place in a bedroom measuring ten-foot square. And many tactical concepts associated with movement, such as hiding behind obstacles or strafing, would become less interesting if performed in the confines of a water closet. Of course, when riding this train of thought we run the risk of derailing into lands having little to do with interactive fiction, but perhaps this is only because it hasn't been explored enough. Who's to say we couldn't have a game that is predominately text-based taking place in a world subdivided into levels? The roguelike genre points in this direction, although as a rule the fictional aspects of these games are suppressed in favor of the very strategy and tactics that levels encourage.

Of course we can go still larger and begin talking about fiefdoms or even entire worlds as units of division. I am sure there is plenty of IF yet to be written that could employ such vehicles to put us in the shoes of great generals or commanders. To convey the myriad decisions and calculations of these characters in the context of a rooms-based structure would prove inelegant at best.

Conclusions

In this discussion we have considered the room as the predominant metaphor for organizing works of interactive fiction. It is a metaphor with a lot of mileage left in it, and indeed it is at the core of all modern IF authoring systems. It is also flexible, and to date there are many games that intentionally ignore or misuse this metaphor so as to do something different. One-room games are an excellent example of this.

But we have also considered alternatives, not so much to deride the room metaphor as to appreciate it, to help us see how it influences the design and implementation of IF today. We have examined the use of more abstract metaphors such as chapters and scenes, and explored the use of temporal mechanisms to subdivide a work into its logical building blocks. We have also considered the crowdedness of a room, and how it leads to an intimacy that can be at odds with settings of a much larger scale. This crowdedness further tempts us to consider ourselves, the players of these games, as being locked into the role of the protagonist. But perhaps this, too, is a metaphor well worth reconsidering?

Nathan D. Jerpe is an indie-games developer based out of Atlanta, GA. He is the author of Legerdemain, a surreal fantasy-role playing game available for download at http://roguelikefiction.com. He is an ardent admirer of roguelikes as well as IF and experimental fiction, and enjoys exploring those wonderful little corners where these mediums intersect.