Il tuo mondo non è il tuo romanzo: una verifica della realtà sul worldbuilding

I once spent seven months building a world for a novel I never wrote.

I had three currencies, each with exchange rates that shifted based on proximity to the capital. I had a calendar with thirteen months, each named after a dead god, and I had worked out which festivals fell on which days and what food was eaten at each. I had a writing system -- not just a fictional alphabet, but a history of the alphabet, including the vowel shift that had happened around what would have been my novel's equivalent of the fifth century. I had a magic system with seven orders, each with a discipline, a color, an organ of the body it was associated with, and a list of its most famous practitioners going back four hundred years.

What I did not have was a protagonist.

I had a notion of a protagonist. She was a woman in her thirties, from the southern province, who was in some sense of the word on the run. But every time I sat down to write her first scene, I found I needed to know one more thing about the world first. What was the trade route between her village and the capital? Who controlled the road tax? What was the local attitude toward outsiders from the eastern kingdom? I would open a fresh document, write three paragraphs of setup, realize I did not yet know what the roadside inns looked like in that region, and return to my worldbuilding file to fix the oversight. Seven months of this. Three currencies. A dead-god calendar. Not a single scene of fiction.

The novel never happened. What I had was not a novel in progress. It was a very elaborate piece of procrastination wearing a novel's coat.

This is the part of worldbuilding nobody warns you about: it feels exactly like writing. You open a document. You type. Words accumulate. There is research involved. There are creative decisions to be made. By every external measure, you are working on your book. But you are not writing your book. You are building infrastructure for a book you have not yet committed to writing, and the longer you build, the more the infrastructure becomes the point.

I want to talk about why this happens, why the standard advice about worldbuilding is half-wrong, and what worldbuilding is actually for -- because the answer is more useful and more specific than "creating an immersive world" or "making your setting feel real." Worldbuilding has a job. Most of the worldbuilding I have ever done, including a lot that was fun to do, did not do that job. Figuring out the difference is, I think, the most important craft question a speculative-fiction writer can ask themselves.

The Tolkien Trap

Every fantasy writer has, at some point, held up Tolkien as their model and their permission slip. He invented two full languages. He wrote genealogies going back to the creation of the world. He sketched maps in ink, rewrote them, redrew them, color-coded them. The appendices to The Return of the King are longer than some novels. If Tolkien could build that much world, surely I can spend another week on currency exchange rates.

Here is the part almost every worldbuilder misses about Tolkien: he was not a novelist who did worldbuilding on the side. He was a philologist who did novel writing on the side. The languages came first -- Quenya and Sindarin were mostly complete decades before Frodo was named -- and Tolkien has said, explicitly and repeatedly, that the stories were invented to give the languages somewhere to live. "The invention of languages is the foundation," he wrote in a letter in 1955. "The 'stories' were made rather to provide a world for the languages than the reverse."

This is an almost unheard-of motivation for writing a novel. Tolkien was not procrastinating on The Lord of the Rings by inventing Elvish. He was writing The Lord of the Rings in order to justify the fact that he had spent half his life inventing Elvish. The worldbuilding was not scaffolding for the novel. The novel was scaffolding for the worldbuilding.

If that is your situation -- if you are a linguist who enjoys constructing languages and has decided to write a novel to house them -- then by all means, build endlessly. That is literally what your book is for. But most of us are not in that situation. Most of us are people who had a story idea, and we are reaching for Tolkien as a reason to keep building a world instead of writing the story. We are misreading his example as a rule when it was actually a radically unusual personal project with a radically unusual outcome.

It is instructive to look at who else in fantasy has produced transcendent work with fundamentally less worldbuilding. Ursula K. Le Guin's Earthsea has no exhaustive appendices. The magic system is small, local, elegant: you learn the true names of things, and knowing the true name gives you power. That is almost the whole system. She does not explain where the names come from, how the language of the dragons relates to the language of wizards in rigorous etymological detail, or what the GDP of the island of Gont looks like. She gives us what the story needs, and not a kilobyte more. And A Wizard of Earthsea is, in my view, a more psychologically complete book than The Lord of the Rings -- not in spite of the thinner world, but because the world does not compete with the character for the reader's attention.

Gene Wolfe's Urth is similarly lean at the surface and bottomless underneath. He will use a word for a weapon or a plant or a governmental office without explaining it, and the reader gradually assembles meaning from context, because Wolfe trusts his world to feel deep without having to stop and prove its depth. He is not showing you the iceberg. He is letting you feel its weight through the way the narrator steps around it.

The Tolkien Trap is the belief that Tolkien's worldbuilding produced his novel. It did not. His novel produced in spite of his worldbuilding, with the help of a preexisting, decades-long linguistic obsession that almost no writer working today shares. The lesson from Tolkien is not "build exhaustively." The lesson is "build what your obsession demands, and then commit to the story that uses it." Most of us do not have his obsession. We should not be imitating his volume.

Worldbuilding Has One Job

Here is the reframe that took me the longest to accept, because it is unromantic and it punctures a lot of the pleasure of the practice: worldbuilding exists to create pressure on your characters. That is its one job. Everything else is decoration.

A fictional world is not interesting because it is detailed. It is interesting because the rules of that world make specific demands on the people living inside it -- demands that would not exist in our world, or that exist in our world in milder forms and are intensified by the fictional transposition. The world is a pressure system. Your characters are the barometer. Worldbuilding that produces no pressure on characters is ornamentation, and ornamentation is what you notice when you are not engaged with the story.

Consider the ring in The Lord of the Rings. It is a worldbuilding object. It has rules (corrupts all who wield it), history (forged by Sauron, lost in a river, found by Bilbo), properties (invisibility, life extension, attraction to its maker). All of these could be treated as lore -- interesting facts about this fictional object. But the ring only becomes novel-sustaining because of what it does to Frodo. The corruption rule is not abstract. It is a specific, escalating psychological pressure that breaks him by Mount Doom. Without the character pressure, the ring is a trivia entry. With it, it is the engine of a nine-hundred-page narrative.

Now consider a worldbuilding detail that does not do this. The geography of Gondor: its provinces, its population distribution, the specific administrative structure of the Stewards' rule. Tolkien has worked all of this out. It is in the appendices. The novel barely uses it. When Aragorn arrives at Minas Tirith, the reader does not need to know the tax base of Lossarnach or the constitutional question of whether the Steward can dissolve the council. The reader needs to know that Denethor is losing his mind, that his grief is killing him, that the city's faith in its rulers is collapsing. The administrative detail is texture at best, and even most of the texture never makes it onto the page.

Which is fine. It is fine to have worldbuilding that does not appear in the book. Tolkien's background material is famously deep, and some of that depth leaks through into the reader's felt experience of the world. But you should know which of your worldbuilding produces character pressure and which is there for your own amusement, because they are not equivalent. One is load-bearing. The other is not. And if you cannot tell the difference, you will spend seven months working on currency exchange rates and wonder why you do not have a book.

The Three Kinds of Worldbuilding

If you audit any worldbuilding document -- yours, mine, anyone's -- you will find three kinds of material living side by side. They look similar on the page. They feel similar to write. They are not doing the same work.

Load-bearing worldbuilding is any detail the plot cannot survive without. The ring's corrupting influence is load-bearing. The existence of the Shire -- a peaceful rural society whose destruction by industrial war is the story's thematic undercurrent -- is load-bearing. In Dune, the scarcity of water on Arrakis is load-bearing: it shapes every Fremen custom, it is the basis of the planet's value, it drives the political-economic machinery of the entire novel. In The Fifth Season, the cyclical apocalypses that the world is built around are load-bearing: they justify the caste system, they produce the orogenes, they explain every social structure that causes the protagonist's suffering. Remove any of these and the novel does not merely lose flavor. It collapses.

Texture worldbuilding is the sensory, atmospheric layer -- the specific smells, meals, clothing, idioms, superstitions, local customs that make a world feel inhabited rather than diagrammed. The hobbits' second breakfast is texture. The way people in Earthsea call each other by use-names to protect their true names is texture (though it leans toward load-bearing because it reinforces the magic system). Texture is what makes a world breathe. It is often the most pleasurable worldbuilding to do and the most rewarding to read. A novel without texture feels like a stage set. A novel that is all texture feels like a museum.

Vanity worldbuilding is the stuff you do for your own pleasure that does not produce pressure or texture -- or does so at a density the story cannot support. My dead-god calendar was vanity worldbuilding. The festivals were never going to appear in the book. The vowel shift in the writing system was never going to come up. I built them because I liked building them, and because the act of building them gave me the same dopaminergic satisfaction as writing would have, without any of the risk of writing badly.

Vanity worldbuilding is not evil. It is often necessary for your own creative immersion -- some writers need to know the full iceberg, even if only the tip is visible, because that knowledge affects how they write the tip. But vanity work masquerades as load-bearing work. It hides in your spreadsheet. It eats your writing time. And it is very difficult to tell, while you are inside it, whether the five-page document you are working on is texture you will eventually draw from or a decorative hobby that will never touch the prose.

The distinction matters most when you are deciding what to do next. A writing session should not be split evenly between "develop the third economic tier of the merchant guilds" and "write chapter four." The economic tier is vanity until a character in chapter four is personally crushed by it -- at which point it becomes load-bearing, and you do it then.

The Chekhov Filter

There is a test I run on every piece of worldbuilding before I commit any more time to it. I call it the Chekhov Filter, after Chekhov's famous dictum about the gun on the wall: if a rifle hangs above the mantelpiece in the first act, it must be fired by the third. The rule is usually applied to prose -- do not introduce objects that do not matter -- but it works just as well on worldbuilding.

Take any worldbuilding detail you have committed to. The currency system. The religion. The magic's cost structure. The succession law. Then ask: what scene does this detail fire in? Which chapter contains the moment where this thing, specifically, causes a character to make a choice they would not otherwise make? If you cannot point to that scene -- not "the whole book benefits from the ambiance" but a specific moment of decision or revelation -- then the detail is not load-bearing. It is either texture (fine, keep it light) or vanity (cut it, at least from the page).

When I apply this test to my own worldbuilding, I am routinely embarrassed by how much of it fails. I built a political conspiracy with four factions and twelve named conspirators for a novel in which, as it turned out, only three of the conspirators ever appeared onstage. The other nine were my dopamine bath. I had a magic system with seven schools, but the protagonist only practiced two of them, and only one of those mattered to the plot. Five of the schools were vanity, disguised as rigor. The moment I forced myself to ask what scene each school fired in, the answer for most of them was none, and I cut them out of the worldbuilding document entirely. The novel got shorter, faster, and better for it.

The Chekhov Filter is uncomfortable to apply because it feels reductive. Your world, you want to believe, is more than the sum of the scenes that use it. And maybe it is, for you, as a private aesthetic artifact. But your reader does not experience your world in the abstract. They experience it one scene at a time. Details that never reach a scene might as well not exist as far as the reader is concerned, which means the time you spent on them was, with respect to the novel, pure overhead.

There is a softer version of the filter for texture work: will this detail color a scene, even briefly? A local superstition that appears in a single line of dialogue is doing its job -- it signals depth without demanding page-time. A three-page document explaining the folklore of the southern provinces, of which only one sentence ever appears in the novel, has returned 0.3% of its effort. That is a terrible return. You would have been better off spending the remaining time writing the actual scene the superstition colors.

Different Writers, Different Systems

Because worldbuilding has one job -- producing character pressure -- writers who understand that job end up taking very different-looking approaches, tailored to the specific pressure each of their stories needs. There is no single correct method. But there is a pattern in the methods of writers who finish novels, versus those who build endlessly and publish rarely.

Brandon Sanderson, in interviews and his YouTube lectures, has articulated what he calls the Laws of Magic -- the most widely circulated of which states that the ability of magic to resolve problems is directly proportional to how well the reader understands it. He writes "hard magic" systems: rigorously rule-bound, taught to the reader early, used to set up precise plot solutions later. This approach requires deep worldbuilding up front, because the climax of the novel depends on the reader knowing the rules well enough to see the solution coming from the right distance. If you are writing a Sanderson-style hard magic novel, your extensive magic system is load-bearing almost by definition.

But -- and this is the part that usually gets missed when Sanderson's laws get quoted -- he also advises building the minimum system required to produce the plot-solving effects you need. He does not encourage writers to build seven schools if four will do. The rigor is in the service of the climax, not of the worldbuilding document. Many Sanderson imitators keep the rigor and lose the ruthlessness, producing five-hundred-page magic systems attached to two-hundred-page novels.

N.K. Jemisin, in The Broken Earth trilogy, takes almost the opposite approach at the prose level -- she withholds lore aggressively, introduces terms without defining them, and trusts the reader to build meaning cumulatively. But underneath, the worldbuilding is doing Sanderson-level plot work: the orogenic abilities, the node maintainers, the stone eaters, the cyclical Seasons are all load-bearing to a degree that is astonishing on rereading. She just refuses to lecture the reader. The iceberg is massive; the visible tip is jagged and strange. The effect is immersive because the reader is doing real interpretive work to assemble the world.

China Miéville tends to build outward from a single strong central premise -- the geographical overlap of two cities in The City and the City, the insectile-human hybrid society in Perdido Street Station -- and then let the consequences cascade. He is less interested in making a complete world than in fully exploring the ramifications of one weird rule. This is, in my view, an underused approach in fantasy. Many novice worldbuilders try to construct the whole world at once, when the more productive move is to find one generative weirdness and interrogate it until it has produced enough texture to carry the book.

Susanna Clarke, in Jonathan Strange & Mr Norrell, builds her world through apparatus: faux scholarly footnotes, references to books that do not exist, an entire pseudo-academic tradition of English magic. This is worldbuilding as performance. The world feels real because the footnotes behave like real footnotes, not because any particular fact about the magic is explained rigorously. Clarke's approach suits a specific register -- the nineteenth-century pastiche -- but the general principle is powerful: you can build a world more efficiently by mimicking the textures of real scholarship, reportage, oral history, or archaeological fragment than by explicating your fictional facts plainly.

What all four of these writers share is that their worldbuilding is tightly bound to the structural needs of their specific books. Sanderson's rigor serves his climactic twist mechanics. Jemisin's withholding serves her thematic argument about trauma and unreliable narration. Miéville's cascade serves his premise-driven structure. Clarke's apparatus serves her register. None of them is doing worldbuilding for the sake of worldbuilding. Every choice is a choice in service of the story. That is the signal you are looking for in your own process: is this choice I am about to make in service of a story I am writing, or is it in service of a world I am amassing?

The Magic System Problem

Magic systems deserve their own section because they are where novels go to die.

There is a specific failure mode I have seen over and over in unfinished fantasy drafts, including my own. A writer gets interested in magic-system design and spends months on rules: what the magic can do, what it costs, what the exceptions are, what the historical schools were, who the most famous wielders were, how magic interacts with technology, what the metaphysical basis is, what lies the orthodox establishment tells about its own powers. Each decision generates three more questions. The system becomes more rigorous by the week. And the novel, which this system was meant to serve, sits untouched.

The reason magic systems swallow novels is that they are uniquely frictionless to work on. Every worldbuilding decision in a magic system is self-contained -- you can design the cost of fire magic without knowing anything about the protagonist -- and every decision produces immediate narrative pleasure in your head (you can imagine scenes using the rule you just wrote). Characters, by contrast, generate friction. They have interiority you have to feel your way into. They resist your plans. They are harder.

So writers do the frictionless work, which looks like writing, and avoid the frictional work, which is the writing. The magic system grows. The novel does not.

The fix, in my experience, is to flip the usual order. Instead of designing the magic system and then writing a character who uses it, pick the character first and build only the parts of the magic system they personally interact with. If your protagonist is a fire mage who has never met a necromancer, you do not need to work out necromancy. You need to work out what it feels like in her body to summon a flame, what the cost costs her specifically, what her relationship to her own power looks like at the psychological level. That is the worldbuilding that will produce pressure on her arc. The seven schools, the ancient orders, the continental history of magical regulation -- that is all for a different book, or for the sequel, or for never.

One way to think about this: a magic system does not need to be complete. It needs to be load-bearing for this protagonist. If a second novel in the same world follows a different protagonist, you can extend the system then, in the direction that new protagonist's arc demands. You do not have to front-load every rule in the first book. Most readers who bounce off hard-magic fantasy are bouncing off front-loading.

Worldbuilding Is Character (Said Another Way)

If worldbuilding's job is to produce character pressure, then every worldbuilding choice you make is, in a disguised form, a character choice. The economic system is interesting to the extent that someone in your novel is personally crushed or enriched by it. The religion is interesting to the extent that someone believes it, disbelieves it, or is torn between the two. The magic is interesting to the extent that someone pays a cost for it. Every rule of your world is someone's prison. That is the perspective that makes worldbuilding a live activity rather than a dead one.

Try this exercise. Take any rule you have written into your world -- say, "fire magic requires the mage to give up something they love" -- and ask: who in this book is locked inside this rule in a way that defines their life? If the answer is vague or abstract -- "well, anyone who uses fire magic" -- then the rule is still a worldbuilding diagram, not a character engine. Keep drilling. Whose life is constricted by the cost? What did they lose? How have they adjusted? What decisions are they now unable to make?

When you have answered those questions, you are no longer building a world. You are building a protagonist. The magic system has collapsed into a specific person's specific psychology, which is exactly where you want it, because that specificity is what the reader will remember. The reader does not remember "magic in this novel had a cost." The reader remembers the scene where the fire mage, now in her thirties, touches the hand of the person she loves and flinches -- because the cost she paid fifteen years ago was the ability to feel other people's warmth. That is the rule made flesh. That is the worldbuilding earning its place.

Family dynamics work the same way. You can chart out the royal lineage of a fictional kingdom in full -- six generations, three civil wars, the bastard son who caused the schism of 1384 -- and none of it does anything for the reader if the protagonist does not feel the specific weight of being, for example, the third son of a third son in a culture where inheritance passes by a particular rule. A family tree is not a document. It is a pressure structure, and every branch is squeezing or cushioning someone specific.

Research Is Worldbuilding, and Research Also Eats Novels

Everything in this article applies to historical and contemporary fiction too, even though the "worldbuilding" is called research in those traditions. The mechanics are identical. You are accumulating facts about a world in order to produce pressure on characters, and the same trap applies: facts that do not produce pressure are vanity, facts that enrich scene-level texture are useful, facts that a character would have to personally reckon with in the plot are load-bearing.

Hilary Mantel's research for the Cromwell trilogy was legendary in its depth. But if you read any interview with her, she is very clear about what the research was for: to know what Cromwell would have known, which is a different question from knowing what a Tudor-era historian would know. The research was an act of mental inhabitation, not encyclopedia-building. Mantel did not put everything she learned into the novels. She put what Cromwell would have heard, noticed, thought, feared -- which is a specific, narrow slice of the sixteenth century, the slice that produces pressure on her protagonist.

I have watched historical novelists get bogged down in research for a decade without producing a book because they kept widening their frame. Instead of "what did a merchant in 1640s Amsterdam know about his own life," they were researching the entire Dutch Golden Age, the economics of the tulip craze, the theology of the Synod of Dort. All interesting. None of it pressure-bearing unless a specific character in a specific scene has to personally deal with it. The fix is the same as for fantasy: pick your protagonist, research only what presses on them, and go write.

When Building Is Avoiding

Let me be direct about the psychology for a moment, because I think it is the real subject of this article.

Writing fiction is, at the core, a sustained exercise in facing what you do not know. You do not know how your protagonist will respond to the inciting incident until you have written her. You do not know whether the scene will work until you draft it. You do not know whether the book is good until you finish it. Every day at the desk is a day of deciding things under uncertainty, with stakes -- because a bad scene takes days to rewrite, a bad chapter can require structural surgery, a bad book can represent years lost.

Worldbuilding feels like writing but is actually, psychologically, much closer to reading. You are accumulating information about a place. You are making decisions, but the decisions are low-stakes -- if you get the currency name wrong, you change it later. If the chapter does not work, you rewrite it fifteen times and it still does not. Worldbuilding is therefore a way to feel productive while avoiding the thing that actually scares you.

I know this because I watched myself do it for seven months. And I have watched dozens of other writers do it for longer. The telltale sign is when a worldbuilding session ends and you feel satisfied and relieved, rather than agitated and eager to get back to the manuscript. Satisfaction after worldbuilding is suspicious. The emotion you want at the end of a good writing session is itch -- the sense that you have barely scratched what the scene is doing, that something remains unresolved, that you cannot wait to go back tomorrow. Satisfaction means you were not risking anything. And if you were not risking anything, you were probably not writing.

This is not a call to abandon worldbuilding. Worldbuilding is a legitimate creative practice. Some writers do it for its own sake and publish the results as gazetteers and companion volumes, which can be great. But if your goal is to finish a novel, and you have been worldbuilding for months without drafting chapters, it is worth asking -- honestly, with the defenses down -- whether the building has become a way of not writing. Usually it has. I have never met a working novelist who regretted writing too early. I have met many who regretted building too much.

A Worldbuilding Audit

If you suspect, as I suspected, that your worldbuilding has started eating your novel, here is the audit I run on my own projects every few months. It is brutal, and it is worth it.

1. List every worldbuilding document you have. All of them. The magic system, the calendar, the map notes, the language glossary, the character histories, the location writeups. Be honest about volume -- not "a few pages" but actual word counts.
2. Next to each document, name the scene it fires in. A specific chapter. A specific moment where a decision is made, a revelation lands, or a consequence hits. If you cannot name one, flag the document.
3. For flagged documents, decide: texture or vanity? If the document contributes mood, idiom, or atmosphere that will color the prose even without appearing in plot -- keep it, but acknowledge it is supplementary. If it is neither firing in a scene nor enriching texture, it is vanity. Archive it.
4. Count the word ratio. Add up your worldbuilding word count and your actual manuscript word count. If worldbuilding exceeds manuscript by a significant multiple, you have a problem. There is no magic ratio, but a ten-to-one worldbuilding-to-manuscript ratio is almost always a diagnosis of avoidance.
5. Pick the next worldbuilding task by scene demand, not curiosity. Instead of "I want to work out the religious schism of the eastern kingdom," ask: what scene am I writing next, and what worldbuilding does that scene need? Build only what the scene requires. If the schism is not in the next scene, park it.

The last point is the hardest to apply, because it fights against the natural rhythm of worldbuilding, which is to follow your curiosity outward. Curiosity is a good servant and a terrible master. It will lead you to build every interesting thing you can imagine. Your novel needs only the things it needs. The audit is there to keep reminding you which is which.

What Worldbuilding Does That Nothing Else Can

I have spent most of this article warning against worldbuilding, which is only half the truth. The other half is that worldbuilding, done right, is one of the most powerful tools in fiction. When a world is load-bearing, it can carry themes that would feel preachy if stated directly. It can show how a society shapes the psychology of the people living in it. It can produce the specific kind of tragic inevitability that only comes from rules no character can change -- the sense, in A Song of Ice and Fire, that winter is coming; the sense, in Nineteen Eighty-Four, that Big Brother will win; the sense, in The Left Hand of Darkness, that gender as we know it is a contingent arrangement and other arrangements are possible. These effects cannot be achieved by character alone. They require the world as a live force.

So the goal is not less worldbuilding. The goal is better-targeted worldbuilding: less volume, more pressure, tighter coupling between every rule and at least one character whose life is squeezed by it. A lean, load-bearing world beats a lush, decorative one every time, because the lean world actually does something inside the novel, and the decorative one just sits there while the reader tries to figure out why they are losing interest.

My failed novel -- the one with three currencies and the dead-god calendar -- taught me this the hard way. I still have those notes somewhere. Maybe one day I will use them. But the lesson was not that I built too little. I had built enough world to fuel three novels. The lesson was that I had built in the wrong direction: outward, encyclopedically, for my own pleasure, when I should have been building inward, toward a character, from the specific pressure points her story required. The world I constructed could have hosted any number of novels. It ended up hosting none, because I never wrote the one character whose life it was meant to crush.

Your world is not your novel. Your novel is one woman, one afternoon, one choice -- made inside a world you have built only in the exact shape required to make that choice impossible and necessary and permanent. Build only that world. Write only that afternoon. Everything else can come later, in the next book, if it earns its place. It almost always doesn't.

Plotiar is an all-in-one workspace for writers who build worlds -- with maps, calendars, family trees, a Lore knowledge base your editor reads as you write, and the chapters those worlds serve, all in a single project. If you are ready to stop building and start writing, start a project for free and see what happens when the world and the story finally share a home.