I profiled Chris Crawford a couple of weeks ago, so that I could discuss his work in future. This post is my first go at it.

We’ll get to this in due course… Source

We’ll get to this in due course… Source

All you really need to know to start with is that Crawford’s an expert in videogame design, who has committed his career to interactive storytelling. His vision of interactive stories has not been realised. He demands a great deal of interactive stories — far more than I’ve been doing in all my discussions so far.

Crawford specifically dismisses a lot of the candidates that I would bring into the interactive storytelling discussion:

We see the gaming design community continue to spin its wheels with the “interleaved story” ploy. This is an old wheeze in which the game alternates between some non-interactive storytelling and some non-dramatic interactivity. . . . This kind of thing has been going on since at least 1990 with Wing Commander, and the only progress I have seen is in the smoothness of visual integration of the two sides of the game.
Chris Crawford in his blog post “Is Interactive Storytelling Stuck?

I have let interleaving into my discussions as a legitimate interactive storytelling technique. I base that approach on an analysis of storytelling into two broad parts: narrative and engagement. Those parts are a little like “the story” on the one hand and “the way it is told” on the other. The storyteller sitting at the campfire tells the same story that has been told round a thousand other campfires, but may tell it well or badly. Part of the experience for the audience are the bare facts of what happens in the story — that’s what I’m calling the “narrative” — the rest of their experience is what I’m lumping together under “engagement”.  If a game only offers interactivity in its engagement then that is still part of storytelling that is interactive — it’s just not the narrative. That’s been my stance on the matter.

For example, a game can use cutscenes to drive narrative efficiently, then use interactive segments to explore the themes of the narrative in a non-linear way, restricted by rules relevant to the story, but not strongly directed through particular actions. These don’t have to be obscure, experimental, bookish indie games: my go-to example is Rez. I described Braid’s storytelling in a similar light, too.

But Crawford demands more for his vision of interactive storytelling.


So what does he want?

It seems that what Crawford’s asking for is interactive narrative (or at least includes interactive narrative). That is to say, he wants all aspects of the story to be engaged with by and dynamically respond to the player — not just some wishy-washy “engagement” with “themes” in an abstract sort of way while the story waits in the wings for the next cutscene.

This has been his life’s work, so I’m convinced he knows what he’s talking about — far more than I know what either of us is talking about. Although I surely won’t do Crawford’s project justice here, there are two interesting parts of the surface I want to try my best to scratch. Here are my questions:

  1. Is interactive narrative an oxymoron?
  2. Does interactive storytelling boil down to just any normal computer game?

Is interactive narrative an oxymoron?

As a contributor to Crawford’s Phrontisterion forum, Stede Troisi, put it in June this year, “Is interactive storytelling even storytelling?” Crawford wrote a response to that post and a very illuminating discussion flowed out of it. I’ll borrow the flow of that conversation to take this post forward in my terms.

To cut the whole thing unduly short, we have a challenge to the notion of interactive narrative and a response from Crawford. The challenge is that narrative is controlled by an author, but interactivity violates that authorial control. There’s a tension and it looks problematic.

In case that idea isn’t intuitive to you, I defined two problems for interactive narrative in my post “Two Problems for Interactive Stories”. For now, I’ll assume we can all at least see the motivation for the challenge.

Crawford begins his response to the issue with a note of advice:

First off, let’s not get caught in this trap:
1. Take an existing non-interactive story.
2. Add interactivity.
3. Is the result any better than the original?
That’s rather like saying that a hamburger with ice cream topping isn’t as good as a hamburger with ketchup.
Chris Crawford commenting on his post

Eventually, discussion leads to Crawford’s conclusion that:

“Interactive storytelling is not the same as stories.” […] “in a conventional game with story, the story is fixed or slightly variable and is bolted on to the game, rather than being the focus of play.”
Crawford in the same comment

What I get from all this is that Crawford’s interactive story — as opposed to a “game with story” — permits and centres on interaction with the stuff of the story itself. In my terms, this means that the narrative — the part of the storytelling experience that is usually non-interactive in the interleaving approach — should be interactive. We have to abandon our preconceptions of narratives that come from non-interactive instances, and be ready to imagine something new, a “story world”, defined by the author, where stories come into existence through interaction.


Does interactive storytelling boil down to just any normal computer game?

So if interaction with narrative must be dynamic interaction in a regulated world, and only the rules and regulations of the world are set by the author, and if the interaction is “play”, have we just got… a computer game?

Crawford’s interlocutor on Phrontisterion puts it neatly:

“What you are talking about is more like the scenarios in most strategy games. The author provides structure, not storytelling. The player provides the details. When the game is over you have a history, not necessarily a story.”
Stede Troisi replying to Crawford’s post

I mean, in Mario there are rules to the world and I interact with the world and am regulated by those rules as I explore it and find a path through it. Is that a story? How about in a game like Journey, where we have multiple characters and meaningful interactions and, again, using the rules of the scenario, a sequence of events emerges as I play?

I don’t know the answer to this question, but again, we need to drop preconceptions and start from scratch with our impression of what the story world is. Forget novels; forget campfire stories; forget (especially) existing videogames. The one analogy that I’ve seen Crawford draw on to help us — in the same discussion — is the soap opera:

“I’ll offer a simple example of a story world that is not interactive but could be made interactive: the world of a soap opera. The basic elements are in place and every day the scriptwriters put the elements together in new ways. Although we could never make a story world as rich as that used in professional soap operas, we could certainly make something with that kind of flavor.”

Crawford continuing the same comment

As far as I can understand it, the story world is a set of actors and stages, with tendencies to act in certain ways. Furthermore, in the pataphysics of the soap opera, there are no non-dramatic events. Harold Bishop never just cleaned his glasses, scratched his arse, and decided to watch television, unless it was some ironic prelude to a momentous happening. The story world is a world configured to generate stories. It’s not a case of “Here are some rules, do as you will”. Instead, do as you will and something significant is going to happen.


I’m confused

Perhaps we need to look at what Crawford has actually been working on. He’s been developing tools to generate on-the-fly story interactions for decades. Here’s a snippet of the successful “first flight” of the story engine for his latest game Siboot:


If anything, I’m more confused

I think Crawford’s idea of an interactive story can be got at by thinking of stories written in text on a page. Now imagine a system that takes text input and outputs text on that basis. Not as a chat bot, and not even as conversing with a real human. It’s not about having a conversation in the game world using language, and it’s not just a display of language parsing, or an open-world, you-can-do-anything technical feat. It’s about the game world being composed only of language, and the player therefore interacting by manipulating the very substance of the game world itself!

If you are in a world of words, and you can spin your own words fluently and quickly and instinctively, and you can interpret words rapidly, deeply, and whilst imagining every implication of them, you have a very special kind of engagement with the world.

So is he making text games? Is it like automatic D&D? No, not really (but I do wonder if storytron isn’t a little bit like a robot DM).

Let’s get more specific. Crawford has noted that the designer of an interactive story must choose between textual and iconic language, each approach having advantages and challenges.

In Siboot, he is using an iconic system, SympolTalk. It consists of “2D sentences”, which look like this:

And consider along with that this abstract for Siboot, the game that will use SympolTalk:

“You are an acolyte on Kira, a moon of the planet Lamina. […] The colonists, thrown onto their own resources, somehow survived under the leadership of Siboot, an extraordinary person who had discovered eeyal, an ESP-type language that permitted the seven species to communicate directly rather than in their separate languages. Over the years, the colonists had learned how to use eeyal, but it was a slow process, and some people picked it up more readily than others. The very best young eeyal-speaker of each species was designated its acolyte, who would become a candidate for leadership when the current leader died.”
Description of the context of Siboot, on Crawford’s blog


OK, I’m ready. Blow my mind.

So the game is about interactions within a non-verbal language. The game consists only of interactions within a non-verbal language. What we have here is an outrageous similitude of system and metaphor. The system and the metaphor are identical, in fact. There is no room for narrative to happen only in the metaphor — such as in a cut scene — and there is no room for interactivity only in the system and away from narrative progression. It’s all of one substance, it’s all story, and it’s all interactive!

Is this what interactive storytelling is? Have I managed even a half-way decent rendition of what Crawford is trying to do? Beats me, but it’s amazing to think about, and if it is what Crawford’s doing then it certainly looks like a life’s work.