Space, Time, Rules, and Rooms: How Videogames Create Narrative Out of Maps

An update: I recently spoke about these ideas at Narrascope 2020. You can watch that talk here.

Vladimir Propp was born in St. Petersburg in 1895. He spent his life as a student and instructor, first of languages and later of folklore. But his best-known contribution to the world is his first book, 1928’s Morphology of the Folktale. In the book, he took the linguistic concept of morphemes — analyzable linguistic units that cannot be divided into smaller parts, that have a meaning on their own (the words “dog” or “man”, for example) — and applied it to narrative.


After graduating university he had dedicated himself to the study of ethnographer Aleksander Afanas’ev’s collection of 600 Slavic folktales, hoping to fill a gap in Russian scholarship, as at the time the field was “totally neglected.” In a series of tales about the “persecuted stepdaughter”, he noticed an interesting fact: the stories followed the same pattern, simply swapping out certain characters. In “Morozko”, a stepmother sends the stepdaughter into the woods to freeze to death, where the monster Morozko threatens to eat her, but she charms him into letting her go. In another tale, a lesij (wood goblin) performs the same task. In yet another, a bear, and so on. “But surely,” said Propp, “they are the same tale!”

Thus were born the 31 famous narratemes. Propp analyzed every folk story in Afanas’ev’s collection from solely the actions of the characters, ignoring specifics, and found that you could rebuild every one of them through combination of these 31 general actions. Just a few are listed here, but from evern a brief look it’s easy to see that he was onto something — all of his narratemes can be applied not only to Russian folktales, but to stories told across the world.


Propp had done something unprecedented by developing a coherent structural system of narrative. But for a long time, no one knew it. Morphology of the Folktale remained obscure, both in Propp’s native Russia and abroad, for almost 30 years after being published, until its translation to English in 1958. In 1960, Claude Lévi-Strauss wrote a response to Propp’s work, perfunctorily praising him for his innovations while criticising him for his dedication to empiricism, arguing instead for his own logical mytheme structure, which also focused on singular events or actions but ignored temporality in favor of analysis of structural relations and binary oppositions. After the Lévi-Strauss review, despite its criticisms, Propp’s methods became famous in the west. He was seen as a man decades ahead of his time, who had predicted the direction of his field thirty years in advance. His work influenced Alan Dundes, Roland Barthes, Melville Jacobs and others.

Though Propp’s flavor of structuralism is no longer en vogue, its methods changed the study of narrative forever. Though others had posited types and categories of stories before him, he was the first to develop a full, coherent system of narrative units. Though his work may have been formalist to a fault — embracing the idea of structural explanation of narrative with little interest in the context around those narratives, or even the effects of recombining those structural elements — it’s undeniably important, and a useful framework for understanding how stories form.


Fifteen years after Propp’s work broke through in the West, another world-shattering innovation was taking place relatively unnoticed. Will Crowther, one of the developers of ARPAnet (the original predecessor to the internet), was going through a divorce. He was an avid spelunker and had taken part in the mapping of Mammoth and Flint Ridge cave systems with his wife Pat before their split. Lonely in the absence of Pat and his children, Crowther spent a few weekends putting together a text-based simulation of the process of mapping a cave, while adding in a few fantasy elements inspired by his experiences playing Dungeons and Dragons. He later wrote that he’d built it in part as a way to re-create his caving experiences with his wife, and in part as something fun for his children. He called it Adventure. After writing the game, he shared it with his coworkers and then went on a month’s vacation. During this time, the game spread across the ARPAnet, becoming wildly popular within the community of researchers that inhabited it.


In the game, the player navigates a cave system. They’re forced to engage in the process of physically mapping out the cave in order to complete it, while dodging axe-throwing dwarves and making their way through a series of devilish mazes. The real locations mapped directly to a cave system Crowther and his wife had mapped themselves. The game, however, took off with plenty of people who’d never mapped a cave system and had little interest in being underground. The joy was not in caving, but in puzzling their way through a coherent, but rigid system.

As Crowther would later put it:

“Why did people enjoy it? Because it’s exactly the kind of thing that computer programmers do. They’re struggling with an obstinate system that can do what you want but only if you can figure out the right thing to say to it.”

Like Propp, Crowther didn’t know it yet, but he’d changed the world. Adventure, later adapted by Don Woods into Colossal Cave Adventure, was the first text adventure game ever made. It would pave the way, first for the rise of text adventures and interactive fiction on early, non-graphical home computers, then later for narrative videogames, interactive films, and every other mix of narrative and computation you can think of. All of this can be traced back to a single, lonely man, writing a program in his spare time to connect with his kids and commemorate his marriage.


This may seem to have very little to do with Vladimir Propp at first glance. But if we’re to take videogames as cultural artifacts, as a narrative form in their own right, it’s important to look at where they depart from past forms, and how we can model them the same way Propp and his contemporaries sought to model stories. Adventure certainly seems like a narrative, albeit a sparse one. And even if Adventure is more map than story, its descendants certainly fall squarely into the category of narrative. Even Super Mario follows a sequence of Proppian functions:

  1. Absentation: Princess Peach is gone.
  2. Departure: Mario pursues her.
  3. Receipt of a Magical Agent: Mario puts on the Raccoon suit.

So on and so forth, encountering various toads and goombas, until….

4. Victory: Mario defeats Bowser, Peach returns to the castle.

But there’s a fundamental tension here, and one that starts back at the beginning with Crowther. The structure of an interactive game is not a story: it’s a map. Ever since Crowther decided to relive his times in Colossal Cave with his wife, games are, structurally, more akin to maps than they are stories. Whether you’re writing a game in FORTRAN or dragging and dropping objects in a Unity interface, games analogize to physical space first, and time second. Crowther chose to make something like a map you could explore: very little in his game is time-based (except things like a magic bridge that could be unlocked with the right key), and his rooms are almost always connected.

The only perceptible passage of time is that of the player going through the game, perhaps occasionally picking up an object — the game space itself almost never changes. The structure of a game, as designed by Crowther, has an odd similarity to the structure of narrative as designed by Propp, but with space as its primary axis. Programmers are, in a sense, the ultimate structuralists. The rules that Propp and his contemporaries hoped to impose on narrative are something like the source code of a game, due to the very nature of programming: without coherent rulesets, the game won’t run. But the ruleset that dominates a program like Adventure relies on spatial context, not temporal.

Meanwhile, Propp’s model of narrative is entirely dependent on time. As he notes, no story contains all of his narrative functions, but they will always occur in chronological order. The can be no punishment without a villainy or lacking, no violation of interdiction before interdiction itself. You can’t save Princess Peach before Bowser takes her from the castle (no matter what Lévi-Strauss might say). This happens naturally in storytelling: the story is told one way, the content does not vary. In Propp’s narratemes, location may vary infinitely, but the actions must always occur in order. Story space is not designed to be explored like a map — going back and forth from room to room- but read start to finish. Temporality is embedded within them. So how do games tell stories when they behave like maps?


Simply put, they have to forcibly impose a narrative. Time has to be forced onto space. And while modern-day games use a number of tricks to hide this process, turning to text adventures allows us to make this process clear. The passage of time has to be tied to space, through the physical process (quite literally physical, when you consider the materiality of the processor on which the game runs) of altering the space the game has created. To make this clearer, consider the “room”.

Rooms — which, in later games, develop into levels — are the phonemes of these games, the units that build their narrative. And the only way in which time can pass in these games is through the traversal and alteration of rooms. In Adventure, for example, to continue past the first level you must pass a room with a great crevice in its center, gather a magic rod in a later room, and later return to the room to wave the rod and reveal a magical bridge. This is a normal narrative sequence taken on its own, but what’s interesting is its utter dependence on location: altering the room with the bridge is the only way to continue the story. Going to the room with the rod is the only way to alter the room with the bridge. But, as I’ve said, Crowther’s story is only nominally a story. His is hardly the best example of the forcible imposition of narrative on games.

For a more contemporary narrative, and one that tells more of a story, we can look at Adam Cadre’s 9:05. After exiting your home, in the game, you are given the option to get into your car and drive to work. In order to continue telling the story, once you enter the car you are now in the car, with your options limited to remaining in the car and getting off at one of two exits. This is not how space works. But it is how time works: once you are on your way to work, you are (barring forgetting your laptop, or a family emergency) on your way to work. By closing off the room you were previously in, and creating a “room” that simulates the passage of time (i.e. a drive to work), Cadre imposes narrative on spatial structure.


This is observable in almost all games that follow a pattern of “levels”: whether you’re playing Half-Life or Mario, the passage of time and the arc of a narrative are shown by the limiting of physical space. Each depends on moving around a map, and then altering that physical map, to tell events in sequence. Once you’ve played through a level, you can’t go back (barring a replay, like rereading a book). Levels, like rooms, are used as substitutes for units of time. To impose Propp’s narratemes on a game, they must occur at a specific place, and they must not be revisitable. Once Mario leaves the first level, he can’t go back: he has received a letter telling him to rescue the Princess. The level has performed its function, and must be left in the past. So the developers close off the physical space to move the story forward.

Modern games perform this spatiotemporal jiujitsu much more subtly, but it’s important to analyze the ways in which games structure narrative differently from traditional narrative forms, and note the limitations that imposes on their ability to do things books and stories can’t: allow the user to explore the space in which the story takes place. To truly simulate space and time, time must flow forward while all the space within the simulation remains accessible, changing organically. But, of course, games can’t do this yet — and probably wouldn’t want to.

This remains the fundamental tension at the center of games: finding the balance between an “open world” and telling a story, giving a game structure and giving users a variety of ways to experience the game itself. It’s no coincidence that the games most often praised for their storytelling are the games that limit their users the most: games “on rails”, like the Half-Life series, The Last Of Us, or Firewatch, that give the impression of an interactive movie rather than a game that has truly variable content. The time that passes in these games is irreversibly tied to the space you traverse while playing them. Meanwhile, while “open-world” games like Grand Theft Auto, Fallout, or Red Dead Redemption, have been praised for their central storylines, one of the greatest joys of playing them is the exploration of game space. To enjoy that exploration, you have to abandon the story, and vice versa. Even when games give you the option of both a tight narrative and a simulated world to explore, the player can’t have both at once.

Perhaps the most interesting response to these limitations is the one offered by games like Gone Home or 80 Days: rather than trying to impose the passage of time on space, they use the exploration of space to uncover a story. Gone Home offers a limited worldspace where a story is told through objects left in that space, 80 Days offers a quite literal map for players to explore. But both of these accomplish their narrative goals through limitation: Gone Home by limiting its worldspace to a single house, 80 Days by using a map as its sole interface. The same problem that faces games “on rails” seems to exist here, but as the base assumption that the game world is built on rather than a process that is imposed over the course of a game’s narrative. What makes these games effective is that they know they are maps.


Narrative videogames were born from the work of a single man, fighting against time, trying to recapture a space that was lost to him forever. Ever since then, that fight has continued within the design of game space itself: time vs. space, narrative vs. choice. Despite the structuralism at the heart of both of their models, Propp and Crowther inadvertently designed two systems that could never fit together quite right.

I wrote A Propper Narrative as a response to this tension, and this article in the hopes of giving a little bit of context to the game. While I offer no solution to the problem of narrative structure in games, I hope reading and playing help more people think about the limits and possibilities that videogames (and specifically text adventures) have to offer as a form of narrative. Honestly, just about any videogame with a story, intentionally or not, is a product of this tension — but I’d love it if you played mine anyway. It’s always a work in progress, but you can give it a whirl here.


Propp, Vladimir. Theory and History of Folklore

Propp, Vladimir. Morphology of the Folktale

Propp, Vladimir. The Structural and Historical Study of the Wondertale

Jerz, Dennis. Somewhere Nearby Is Colossal Cave

Dundes, Alan. Binary Opposition In Myth: The Propp/Lévi-Strauss Debate in Retrospect

Written by

Programmer, occasional writer, flaneur

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store