Worlds are story scaffolds
Simple is simple, but lots of simple is complex
Kurt Vonnegut: “Stories have simple shapes.”
One of the most popular is a cup:
We call this story “man-in-hole,” but it needn’t be about a man, and it needn’t be about a hole. Here’s a good way to remember: somebody gets in trouble and gets back out of it.
People love that story. They never get sick of it.
Vonnegut’s eight story shapes hang on two axes: fortune and time. One could argue that these two dimensions are how humans think about their place in the world. At the end, am I better off than when I started?
Good stories are morally simple and temporally bounded, but good worlds are not. Good worlds allow for ambiguity and open-endedness. A single story won’t exhaust a world. In the background there are untold histories, uncharted territories, unidentified creatures. The happy ending of a “man-in-hole” story may be the start of another character’s “from-bad-to-worse” story.
Good worlds are story scaffolds, including the good, bad, and boring ones. Worlds contain way more ”nothing-in-hole” stories than “man-in-hole” ones: a man digs a hole, forgets about it, and nothing falls in it for years and years and years. Maybe one day this hole plays a role in someone else’s story. Maybe not.
To build a world from scratch, the architect cannot simply make things up. They need to know what interesting stories ought to be set in the world, so that it can supply the necessary story elements—chiefly, a definition of fortune.
In adventure stories, fortunes are riches; in romance, it’s love. In a coming-of-age story, it’s identity; in a mystery, it’s truth.
The same principle applies to storytelling in corporate worlds. A great sales quarter ends with a bag of money, while the absence of incidents may define a good one for the security team.
Data’s role is to make these world elements tangible. It gives storytellers the raw ingredients they need to build compelling stories. While it’s all good for a fiction story to lean on imagination, there’s a higher bar for storytellers when it comes to resource allocation.
Imagine a fast fashion startup, ThneedCo, is looking to summarize its first quarter of business. Their CEO is going to have one definition of good fortune: sales. If ThneedCo has done well, its story might follow this pattern:
Good job, ThneedCo team!
But selling all those Thneeds has knock-on effects elsewhere in the world of ThneedCo. Some customers aren’t satisfied with the quality, and they’ve returned their Thneeds. That’s ill fortune, and a concerning story.
There’s another nuance as well. The increased demand and support burden mean that the team has hired more people. On the one hand, that has increased capacity, but it’s also increased coordination costs. In recent weeks, Thneed-veloper velocity has started tanking.
Is the growth net good fortune? It’s ambiguous.
Meanwhile, one consultant is making a fuss about sustainability and the company’s impact on the local environment. For every Thneed, some Gluppity-Glup is created as a by-product, along with a bit of Schloppity-Schlop. But the company hasn’t started collecting data on this—there is too much else going on.
That creates a foggy area in the world. The degree of pollution is unspecified, creating room for disagreement even among good-faith navigators. The uncertainty precludes a good, crisp story from being told.
All of these elements—revenue, returns, velocity, gloppity-glop—are first-class entities in the world. They are something that someone who spends a great deal of time interacting in the world can reasonably expect to encounter, even if they aren’t essential to that person’s main story.
One of the most challenging aspects of world-building is the degree of resolution at which these entities are available. Over-specification of an entity—e.g., “revenue from Thneed #9317430 collected in Whoville”—often prevents good storytelling as much as underspecification.
Take, The Good Samaritan, which is a classic man-in-hole parable. In it, a Jewish man is walking from Jerusalem to Jericho, but he’s attacked by robbers and left for dead on the side of the road. Two Jewish men, both religious officials, pass by the man but decline to help him. Finally, a Samaritan man discovers him, tends to the man’s wounds, and gets him to safety. The moral is that neighborly love is shown through actions, not shared background.
For listeners, the moral lesson hinges on world context: that Jews and Samaritans did not get along at this time. It’s shocking that the injured man’s “good fortune” would come from someone outside his people. There’s a history that enables the story to work.
There’s plenty more history that goes unmentioned in this parable. The storyteller has decided to exclude it, but if this were an artificial world being built, could its builder ignore the plethora of other types of people and their relationships?
Where would you draw the line on which ethnicities to represent? Certainly, the Greeks, Romans, and Syrians are big enough to describe first-century Jerusalem, but what about the Egyptians? The Nabataeans? Is it worth including subdivisions, minor rivalries, and regional differences in the world substrate?
These are questions of resolution, and there is no single answer as to how complex the world should be. However, some principles are worth following.
First, tell a few simple stories. In this world, what will people think of as good fortune? What is bad? If that isn’t clear, it will be hard for the people who want to use the world’s data to tell their stories.
Second, settle on entities before anything else. These entities are the primary ingredients for telling different stories, and most of the interesting dynamics of the world will be in how these entities interact. Think—it’s easier to postulate the existence of a new city than it is to bring in a race of elves.
Lastly, start at low resolution and build in dimensionality. The Good Samaritan could work even without the additional nuance about the Samaritan—i.e., love is shown through action. It wouldn’t have lasted 2000 years, sure, but it works.
As the world grows in size and complexity, small nuances take on oversized roles, and more attention is given to small-group interactions. If a world is well-modeled and generalizable, “drilling down” into these details should come naturally.
But it needs to starts simple: simple stories in simple worlds. Because as more of these stories stack on top of each other, the world becomes crowded, populated, nuanced. Supporting one hundred simple stories told by one hundred different people in a single world?
That’s complex.