Worldbuilding: how to build a fictional world that works

Strong worlds usually rest on a single premise-level difference from our own: the dead remember, the tide never comes in, the empire runs on harvested dreams. Start there, not with the map. One deep change, honestly explored, out-worlds a dozen shallow ones — because your reader learns one rule and then feels every consequence, rather than memorising trivia.

The exploring is done with one question, asked relentlessly: what does it cost, and who pays? If dream-harvest powers the empire, someone farms the sleepers, someone polices the waking, and someone's child sleeps too long. The cost side of your premise is where the plot lives — a difference nobody pays for is decoration.

Whether it is sorcery or a jump-drive, invented power plays the same structural role, and it needs the same three declarations: what it can do, what it cannot do, and what it costs to use. The limits matter more than the powers — a climax is only honest if the reader knows the hero cannot simply cast the bigger spell. Decide the edges early and never cross them; a rule broken for convenience refunds the reader's investment in every rule.

The classic distinction is hard versus soft: hard systems have explicit, learnable rules (the reader can predict what is possible), soft systems keep magic numinous and unexplained. Both work — the craft law, popularised by Brandon Sanderson, is that your ability to *solve* problems with magic is proportional to how well the reader understands it. Soft magic makes wonderful problems and terrible solutions.

A culture built from food, dress and festival names is a costume — pretty, and inert. A culture becomes alive the moment it disagrees with itself: the young who break the fishing taboo, the priests split over the new comet, the merchants who quietly resent the warrior caste they fund. Internal argument is what lets you write members of a culture as people rather than representatives.

The fastest depth-builders: decide what each group believes about the others (prejudice and myth beat demographics), what a person there owes — to family, faith, guild, crown — and what happened a generation ago that no one discusses at dinner. That last one, the recent wound, does more work per sentence than any other worldbuilding fact; it is the question the template asks first.

Build the world your story visits and let the rest stay fog. You need travel times more than coastlines (distance is tension: how long until the army arrives?), and you need the three to five places where scenes actually happen, each with a smell, a sound and a reputation. A vividly rendered inn beats an accurately rendered continent.

History works the same way: two or three ancient events that every culture tells differently, plus the recent wound in living memory. History that no character disputes is a textbook; history they are still fighting about is a plot. If you are building for an epic or a series, this is where the full 22-question bible earns its keep — and where a novel outline keeps the discoveries pointed at a story.

Worldbuilder's disease is real: the world is infinitely extendable and the blank page is not, so the building never quite finishes. The honest test — can you answer the ten sketch questions in the template above? Then you can start the draft, and the world will grow truer through writing than it ever would through planning, because scenes ask questions no questionnaire can predict.

Keep the bible open while you draft and record decisions as you make them — worldbuilding during revision is fact-checking; worldbuilding instead of drafting is procrastination with a map. When the world starts suggesting people, give them pasts with the character backstory generator, and go find out what your premise costs somebody in particular. That is chapter one.