Shared publicly  - 
Narrative isn’t usually content either
Crossposted from

When I said that narrative was not a game mechanic, but rather a form of feedback, I was getting at the core point that chunks of story are generally doled out as a reward for accomplishing a particular task. And games fundamentally, are about completing tasks — reaching for goals, be they self-imposed (as in all the forms of free-form play or paideia, as Caillois put it in Man, Play and Games) or authorially imposed (or ludus). They are about problem-solving in the sense that hey are about cognitively mastering models of varying complexity.

Some replies used the word “content” to describe the role that narrative plays. But I wouldn’t use the word content to describe varying feedback.

In other words, perverse as it may sound, I wouldn’t generally call chunks of story “game content.” But I would sometimes, and I’ll even offer up a game design here that does so.

The usual definition of “content” is “everything that isn’t code or rules,” meaning all the art and voiceovers and quests and whatnot. But that’s not what it means in this context, because we’re embarking on another one of thse annoyingly formalistic exercises here. :)

I have previously described the basic model I use for analyzing games formally as “a game grammar.” This was mostly a conceit for a presentation title, but in point of fact it fits the formal definition of “grammar” moderately well. You see, this model, which I have also termed an atomic model of game design, is concerned exactly with the morphology of games: the structure and form they take. It builds on the seminal work of Chris Crawford, who defined interaction as

"a cyclic process in which two actors alternately listen, think, and speak."

– Chris Crawford in The Art of Interactive Design

The game grammar model works the same way as all interaction does. The chief difference with game interaction is that one of those actors may actually be algorithmic: a computer, or a set of rules and processes. At core, a game is about figuring out the rules and processes that an opponent is using; said opponent might be a computer or a real person, or even the laws of physics and the physical constraints of your own body. Your job is to identify a goal (which might be handed to you by a designer, or might be one you set for yourself) and attempt to arrive at a way of interacting with this system that results in the outcome you want.

When we speak of a game system, that collection of rules is what we mean. Usually a system will be composed of multiple mechanics, each of which is made of up a variety of rules. A system like this has also been termed a “fun molecule,” an “atom” or a “ludeme” by various authors.

A system, though, is sort of like an algorithm, or a printing press. It repeatably performs a process, but given different stuff to work with, you can get a pretty different experience out of it. The term for the “stuff to work with” is content, and most of the time it is effectively “statistical variation.” An enemy with different stats, a level with different placement of platforms.

There is a class of games that focuses on user-generated narratives rather than on authorially imposed ones — you can read about the distinction in a very old talk called “Two Models for Narrative Worlds” I gave at the Annenberg Center at USC. In that talk I made the point that

"These worlds can still tell stories. What we surrender is not narrative, but authorial control."

I coined the terms “impositional space” and “expressive space” to define the ends of this spectrum for myself.

Now, that talk long predates any of the game grammar sort of work. But effectively, my critique of quick-time-events and excess feedback used in narrative-driven games is primarily about impositional spaces, narrative imposed by the author(s) of the game; and it is essentially in a “ludic” context. And several folks took me to task for ignoring the expressive spaces and the spaces that are intended to serve as narrative generators in that critique.

Story, as it happens, has some rules too, largely based on how the brain works. For example, in The Art of Fiction John Gardner has a wonderful example of the ways in which repeated mention of physical objects causes them to become associated with emotions — in effect to become symbols. And then mention of objects associated with those objects does the same. In a sense, thematic freight becomes transitive.

That particular trick is used very very widely in all sorts of media. For example, Ravel’s Bolero has become thoroughly associated with sex thanks to the film 10, and now at this point you can conjure up that association by just playing that music.

Expressive spaces in games rely on this trick extensively. In fact, all forms of post facto storytelling by players do. They ascribe meaning to moments, and then the player builds a narrative arc through their selective memory of events. I often call this mythmaking, and we do it pretty much all the time, without even thinking about it.

In games designed to cause the player to put together stories, such as Sleep is Death, Facade, or Dear Esther, there is a system there, an algorithm — and then there is the statistical variation that is fed into it. And that statistical variation, the content, is actually little symbols and narrative moments, ones that are often impressionistic or disconnected. The “problem” the player faces is that of arranging them into a coherent whole.

The fact that symbols and moments and memories are profoundly intangible things does not mean that they can’t be manipulated in this way; fiction does so readily, as we have seen. From a mechanical point of view, though, they have much in common with the particular hand of cards you have been dealt, or the set of Scrabble tiles on your rack. You end your interaction with the system by making sense of them, which is different from finding a word in the tiles only by a matter of degree. Dear Esther‘s mechanics could be replicated with a different setting and group of symbols — to radically different emotional effect. When analyzed by the game grammar, we’d find two very different experiences to be the same game.

* * *

Let’s consider a thought experiment.

I was once in a discussion with some fellow designers and one of them was playing with the idea of a game about memories. I offered up a design idea whereby there was a map of a house, and there was a deck of cards, each card labelled things like “comfy armchair” and “deep closet” and “empty bookshelf.” The deck was shuffled, and some cards were laid in each room.

Players would then take turns tapping a card and telling a “memory” about that card and its place in that house. That this was the armchair where you remember curling up to read, a memory of safety and comfort; and another player says it was where they found great-grandmother when she finally passed away. All memories must be “true” — meaning, they cannot contradict anything anyone has said. After all stories were told, all the players decide which way they want to remember the armchair from among the stories told, by voting.

The person whose memory was selected keeps the card. At the end of the game, whoever has the most cards wins.

For greater emotional impact, you play this with real family, a real house layout, and real objects from your childhood.

Here we have both emergent consensus narrative and a game system. The memories are actually tokens in the game space — intangible ones, with a lot of emotional weight to them. You can approach the game mechanistically, and strategize. But you can also approach it experientially. It is mostly an expressive space. And ultimately, the real game lies in making sense of your family, its history. It is still pattern-matching, grokking each other and the complex web of relationships and half-truths and biased recollections that make up a family history.

In this game,

- narrative is input — the affordance given to a player, the “move they can make”
- narrative is a resource — accumulated and managed towards a victory condition
- narrative is actually content, user-generated even, providing statistical variation into the system
- narrative is feedback — its accumulation, in the form of individual symbols, is representing the gestalt “game state”

But it’s still not a mechanic. You could in fact replace the memories with differently colored poker chips, and everything would proceed in the same manner. The experience would be substantially different, and the emotional impact far less.

You could also de-game this. Don’t negotiate whose memories win out. Don’t have the rule about non-contradiction. You’d end up with the experience of looking through a photo scrapbook — and likely, you would not tackle the challenge of understanding that the rules push you towards.

This game has never been played. If anyone ever does, let me know what happens.

* * *

In the post title I said that narrative isn’t usually content. This game is an exception, as are the other ones I have cited. Ironically, games where narrative is content actually tend to have very very complex and robust rule systems. Chris Crawford’s Storytron has years of development in it, almost all in the systems design. Facade is an AI wonderment. And even this little non-digital game has as “imported” rules a host of psychology and past family history, rules that are deeply perilous to transgress. (The mere addition of other players always imports complex social rules into a game; in this case, the deeply personal nature of the interaction brings in yet more. “We never talk about her drinking problem” and the like).

Because of this, I have no issue reconciling formalism in examining the “ludology” of games with the “narratological” approach of examining games-as-stories. My issues with small-system-big-feedback games described in the other post have to do with the lack of substantive pattern-learning, the lack of player agency, and thus the lack of the fundamental qualities that games bring to the table. And in that, I include emergent-narrative games and expressive spaces, which I certainly consider games — more complex games, in point of fact, than most games are. So for those who felt I was bashing the entire genre of emergent narrative games, I apologize for the lack of clarity there; that was not at all where I was going with that post.

So where does this all leave authorially imposed story? Primarily in the realm of interactive experience design. Which is a different discipline from “game design” though they have tremendous overlap. I am biased towards our getting game design right, but that does not mean that interactive experience design isn’t a fascinating and deep area in its own right — or that it is unimportant to games. In fact, it’s incredibly important. But that’s a subject for another post someday.
Owen Johnston's profile photo
Add a comment...