A fictional world is potentially infinite in its detail, just like the real world. How do you decide when to say, that’s enough — I’m done worldbuilding?

This question is, among other things, a matter of taste, which means there’s no actual “right answer” apart from the one that suits your preferences. I think it’s pretty obvious from my essays here, my own novels, and the things I enjoy reading that I set the bar on the higher end: I like to feel immersed in a believable world, one that has enough texture for me to sink into the illusion of its reality. Other people would look at books I love and find them tediously over-detailed.

But I’d argue this isn’t purely driven by taste. Or rather, taste sets the upper and lower bounds for your preferences; within the resulting range, there are certain factors you can use to guide your decision-making.

The first is that I think it’s beneficial for a story to have at least a little more worldbuilding than the plot strictly requires. The word “strictly” is key to this equation, though; I am really and truly talking about what the plot absolutely has to mention in order to make the story go. This came to my attention when I tried reading a novel (which, unsurprisingly, I did not finish) that appeared to subscribe to the “only what is required and not an adjective more” school of thinking. Nothing was described unless it was load-bearing — for example, the first mention of anyone’s clothing was when a character showed up wearing the uniform of a dangerous organization. What did that uniform look like? Dunno, because that doesn’t matter. The factual content of “this is why the protagonist knows to be wary of him” was sufficient, and the visuals were treated like unnecessary padding. The story was so bare bones, even a skeleton would eye it askance. And the result was that I had a direct view of the plot machinery operating, without any skin of verisimilitude to make it feel more natural.

Of course, you don’t want to mention that which is entirely irrelevant. In the setting of the Rook and Rose trilogy, when Vraszenian settlements make war on one another, the victorious side raids the labyrinth (temple) of the losing side and steals all the Faces (representations of the benevolent aspects of the deities) as their trophies, leaving behind only the Masks (representations of the malevolent or wrathful aspects). You will find this detail nowhere in the books — admittedly because I thought it up after my co-author and I had finished writing them! But even if I’d come up with that detail sooner, it still wouldn’t be in the trilogy, for the simple reason that nowhere in our story do we have one Vraszenian settlement making war on another. Any reference to that practice would be air-dropped in from the stratosphere, disrupting the story we’re actually telling.

One solution to this, as I’ve mentioned before, is to make your cool idea relevant. We’d have a hard time doing that with internecine warfare, but we could have worked it in as a side note: we do have a conflict between street gangs, and maybe they do something metaphorically similar, which would be reason for someone to mention the larger-scale practice in passing. In this case it’s still a reach, but it serves to illustrate how, if you already have a super-shiny idea, you can look for ways to integrate it with your narrative. From the perspective of “where do you stop worldbuilding,” though, the answer is “before you reach this point, unless the idea comes to you of its own accord” (as this one did). There would be no purpose in me asking “okay, so what does warfare look like in the rest of Vraszan, when city-states or neighboring villages get into conflicts?” when that’s entirely tangential to the actual plot.

You also have to keep an eye on your pacing. Let’s say your protagonist is writing a letter in cipher: should you spend time figuring out what type of cipher their society uses? It’s a relevant question . . . but the details could potentially bog down your scene, stalling the reader with minutiae that distract from the content of the message itself. Personally, I’d be more likely to go in-depth on that question if the character was trying to crack a cipher, because now it’s a challenge they’re trying to overcome — automatically more interesting thanks to the unknown contents of the letter. If you start to research or brainstorm on something, then realize it’s drawing you away from the forward momentum of the story, limit it to a line or two of description at most, or just let the reader supply whatever default lives in their brain.

Finally, is your worldbuilding stopping you from writing the story? It’s one thing if you genuinely need to know something in order to move forward. That happened recently with me getting some distance into the draft of an upcoming novel about a monk going on a pilgrimage, then stopping because I needed to do a lot more development of both my map and my calendar if I didn’t want the pilgrims to be floating in a vague, timeless void. That’s one I maybe should have seen coming and taken care of sooner . . . but there’s a lot of worldbuilding stuff you don’t know you’re going to need until you sit down to write a given scene.

So the notion that you will do all your worldbuilding first and then, when that’s complete, write your story? That’s a trap, one that can keep you forever in the planning stages and never in the execution. Some things you have to know in advance: I couldn’t have started the novel without a basic sense of the religion my monk protagonist follows. It’s entirely legitimate to lay some groundwork before you begin. Much of your setting, though, can and probably should grow with the characters and the narrative, shaping and being shaped by the specifics of the tale. You’ll get a more organic, real-feeling result that way than if you lay down a bunch of shiny ideas in advance and then shoehorn everything in around them.